1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks. This
36 was broken by hardening introduced for Bug 3033.
43 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
44 SMTP connection" log lines.
46 JH/02 Option default value updates:
47 - queue_fast_ramp (main) true (was false)
48 - remote_max_parallel (main) 4 (was 2)
50 JH/03 Cache static regex pattern compilations, for use by ACLs.
52 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
53 Make the rewrite never match and keep the logging. Trust the
54 admin to be using verify=header-syntax (to actually reject the message).
56 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
57 (under Linux) we watch the dir containing the final file; previously
58 it would be the dir with the first symlink. We still do not monitor
61 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
62 dn_expand() is happy to pass them through.
64 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
65 the file with the proof had an unchanged name, the new proof(s) were
66 loaded on top of the old ones (and nover used; the old ones were stapled).
68 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
69 more than one message arrived in a single connection a reference from
70 the earlier message could be re-used. Often a sigsegv resulted.
71 These variables were introduced in Exim 4.87.
72 Debug help from Graeme Fowler.
74 JH/09 Fix ${filter } for conditions that modify $value. Previously the
75 modified version would be used in construction the result, and a memory
78 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
79 Find and fix by Jasen Betts.
81 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
82 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
83 the systemwide configuration to override the Exim config.
85 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
86 API changes in libopendmarc.
88 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
89 pid 1, in the normal "background daemon" mode, having to drop process-
90 group leadership also lost track of needing to create listener sockets.
92 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
93 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
95 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
96 a capture group which obtained no text (eg. "(abc)*" matching zero
97 occurrences) could cause a segfault if the corresponding $<n> was
100 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
101 included a close-brace character (eg. it itself used an expansion) an
104 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
105 starting TLS. Previously it was after, meaning that attackers on such
106 ports had to be screened using the host_reject_connection main config
107 option. The new sequence aligns better with the STARTTLS behaviour, and
108 permits defences against crypto-processing load attacks, even though it
109 is strictly an incompatible change.
110 Also, avoid sending any SMTP fail response for either the connect ACL
111 or host_reject_connection, for TLS-on-connect ports.
113 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
114 Previously this was not permitted, but it makes reasonable sense.
115 While there, restore a restriction on using it from a connect ACL; given
116 the change JH/16 it could only return false (and before 4.91 was not
119 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
120 was exactly sized compared to the log buffer, a crash occurred with the
121 misleading message "bad memory reference; pool not found".
122 Found and traced by Jasen Betts.
124 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
125 dns_again_means_nonexist included an element causing a DNS lookup which
126 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
127 included (though probably not limited to) a process crash from stack
128 memory limit, or from excessive open files. Replace this with a paniclog
129 whine (as this is likely a configuration error), and returning
132 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
133 this always failed, probably leading to the usual downgrade to in-clear
136 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
137 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
138 to not checking that list for these lookups.
140 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
143 JH/23 Fix crash in string expansions. Previously, if an empty variable was
144 immediately followed by an expansion operator, a null-indirection read
145 was done, killing the process.
147 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
148 include an SMTP response string which is longer than that supported
149 by the delivering transport. Alleviate by wrapping such lines before
152 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
153 chars (RFC limit). Previously a limit of 12 items was made, which with
154 a not-impossible References: in the message being bounced could still
155 be over-large and get stopped in the transport.
157 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
158 close. Previously a bare socket close was done.
160 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
163 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
164 message being considered for delivery was already being handled by
165 another process, and having an SMTP connection already open, the function
166 to close it tried to use an uninitialized variable. This would afftect
167 high-volume sites more, especially when running mailing-list-style loads.
168 Pollution of logs was the major effect, as the other process delivered
169 the message. Found and partly investigated by Graeme Fowler.
171 JH/29 Change format of the internal ID used for message identification. The old
172 version only supported 31 bits for a PID element; the new 64 (on systems
173 which can use Base-62 encoding, which is all currently supported ones
174 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
175 and must use Base-36). The new ID is 23 characters rather than 16, and is
176 visible in various places - notably logs, message headers, and spool file
177 names. Various of the ancillary utilities also have to know the format.
178 As well as the expanded PID portion, the sub-second part of the time
179 recorded in the ID is expanded to support finer precision. Theoretically
180 this permits a receive rate from a single comms channel of better than the
182 The major timestamp part of the ID is not changed; at 6 characters it is
183 usable until about year 3700.
184 Updating from previously releases is fully supported: old-format spool
185 files are still usable, and the utilities support both formats. New
186 message will use the new format. The one hints-DB file type which uses
187 message-IDs (the transport wait- DB) will be discarded if an old-format ID
188 is seen; new ones will be built with only new-format IDs.
189 Optionally, a utility can be used to convert spool files from old to new,
190 but this is only an efficiency measure not a requirement for operation
191 Downgrading from new to old requires running a provided utility, having
192 first stopped all operations. This will convert any spool files from new
193 back to old (losing time-precision and PID information) and remove any
194 wait- hints databases.
196 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
197 we treated them as item separators when parsing for a list item, but they
198 need to be protected by the doublequotes. While there, add handling for
201 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
202 Found and fixed by Jasen Betts. No testcase for this as my usual text
203 editor insists on emitting only valid UTF-8.
205 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
206 a null-indirection SIGSEGV for the receive process.
208 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
209 -bh use. Internal checking would spot this and take a panic.
211 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
212 In 4.96 this would expand to empty.
214 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
215 certificate. Find and fix by Andreas Metzler.
217 JH/36 Add ARC info to DMARC hostory records.
219 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
220 or fakedefer. Previously the sender could discover that the message
221 had in fact been accepted.
223 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
224 sequences. Previously the input was not noted as being tainted; notably
225 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
226 bad coding of authenticators.
228 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
229 and ${tr...}. Found and diagnosed by Heiko Schlichting.
231 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
232 could be triggered by externally-supplied input. Found by Trend Micro.
235 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
236 be triggered by externally-controlled input. Found by Trend Micro.
239 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
240 be triggered by externally-controlled input. Found by Trend Micro.
243 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
244 Make the rewrite never match and keep the logging. Trust the
245 admin to be using verify=header-syntax (to actually reject the message).
247 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
250 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
256 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
257 after reception to before a subsequent reception. This should
258 mean slightly faster delivery, and also confirmation of reception
261 JH/02 Move from using the pcre library to pcre2. The former is no longer
262 being developed or supported (by the original developer).
264 JH/03 Constification work in the filters module required a major version
265 bump for the local-scan API. Specifically, the "headers_charset"
266 global which is visible via the API is now const and may therefore
267 not be modified by local-scan code.
269 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
270 sendfile() didi not account for the way the ClamAV driver code called it.
272 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
273 time check was being done for every character; replace that with one
276 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
277 sent was prefixed with a length byte.
279 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
280 RFC 5321. Previously Dovecot (at least) would log errors during
283 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
284 in 4.61 without a reason logged. Binaries will be bigger, which might
285 matter on diskspace-constrained systems, but debug is easier.
287 JH/09 Fix macro-definition during "-be" expansion testing. The move to
288 write-protected store for macros had not accounted for these runtime
289 additions; fix by removing this protection for "-be" mode.
291 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
292 handing out large-numbered file descriptors, violating the usual Unix
293 assumption (and required by Posix) that the lowest possible number will be
294 allocated by the kernel when a new one is needed. In the daemon, and any
295 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
296 are not useable for FD_SET() [and hence select()] and overwrite the stack.
297 Assorted crashes happen.
299 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
300 main-section options or in a connect ACL, the value from the first ever
301 connection was never replaced for subsequent connections. Found by
304 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
305 though only once PCRE2 was introduced: the memory accounting used under
306 debug offset allocations by an int, giving a hard trap in early startup.
307 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
309 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
310 with underbars is given. The write-protection of configuration introduced
311 in 4.95 trapped when normalisation was applied to an option not needing
314 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
316 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
317 failing to close the certificates directory, every hour or any time it
320 JH/16 Debugging initiated by an ACL control now continues through into routing
321 and transport processes. Previously debugging stopped any time Exim
322 re-execs, or for processing a queued message.
324 JH/17 The "expand" debug selector now gives more detail, specifically on the
325 result of expansion operators and items.
327 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
328 bad comparison between the option value and the name of the file to
329 be included was done, and a mismatch was wrongly identified.
330 4.88 to 4.95 are affected.
332 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
334 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
335 supplied as the db file-pair basename. Previously, if a directory
336 path was given, for example via the autoreply "once" option, the DB
337 file.pag and file.dir files would be created in that directory's
340 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
341 "taint" log_selector. These were previously deprecated.
343 JH/22 Fix static address-list lookups to properly return the matched item.
344 Previously only the domain part was returned.
346 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
347 the call into OpenSSL to send a TLS Close was being repeated; this
348 resulted in the library waiting for the peer's Close. If that was never
349 sent we waited forever. Fix by tracking send calls.
351 JH/24 The ${run} expansion item now expands its command string elements after
352 splitting. Previously it was before; the new ordering makes handling
353 zero-length arguments simpler. The old ordering can be obtained by
354 appending a new option "preexpand", after a comma, to the "run".
356 JH/25 Taint-check exec arguments for transport-initiated external processes.
357 Previously, tainted values could be used. This affects "pipe", "lmtp" and
358 "queryprogram" transport, transport-filter, and ETRN commands.
359 The ${run} expansion is also affected: in "preexpand" mode no part of
360 the command line may be tainted, in default mode the executable name
363 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
364 the the facility was not passed across execs, and only the first message
365 passed over a connection could use BDAT; any further ones using DATA.
367 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
368 uses $sending_ip_address and an interface is specified.
369 Previously any use of the local address in the EHLO name disabled
370 PIPECONNECT, the common case being to use the rDNS of it.
372 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
373 resumption. Previously verify failed because no certificate status is
374 passed on the wire for the restarted session. Fix by using the recorded
375 ocsp status of the stored session for the new connection.
377 JH/29 TLS resumption: the key for session lookup in the client now includes
378 more info that a server could potentially use in configuring a TLS
379 session, avoiding oferring mismatching sessions to such a server.
380 Previously only the server IP was used.
382 JH/30 Fix string_copyn() for limit greater than actual string length.
383 Previously the copied amount was the limit, which could result in a
384 overlapping memcpy for newly allocated destination soon after a
385 source string shorter than the limit. Found/investigated by KM.
387 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
388 close; it may be needed for a subsequent connection. This caused a
389 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
391 JH/32 Fix CHUNKING for a second message on a connection when the first was
392 rejected. Previously we did not reset the chunking-offered state, and
393 erroneously rejected the BDAT command. Investigation help from
396 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
397 an empty address. Previously the expansion returned an error.
399 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
400 proxy. Previously these were misparsed, leading to paniclog entries.
406 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
407 related applications. Previously an "H" was used where available info
408 says that "M" should be, so change to match.
410 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
411 as arguments, so an implementation trying to copy these into a local
412 buffer was taking a taint-enforcement trap. Fix by using dynamically
413 created buffers. Similar fix for radius expansion condition.
415 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
416 reasonable, eg. to count headers. Fix by using dynamically created
417 buffers rather than a local. Do similar fixes for ACL actions "dcc",
418 "log_reject_target", "malware" and "spam"; the arguments are expanded
419 so could be handling tainted values.
421 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
422 broken the (no-op) support for this sendmail command. Restore it
423 to doing nothing, silently, and returning good status.
425 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
426 record path was given (or the default used) without a leading directory
427 path, an error occurred on trying to open it. Use the transport's working
430 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
431 transport. Previously it was the name on the DNS A-record; use instead
432 the head of the CNAME chain leading there (if there is one). This seems
433 to align better with RFC 6125.
435 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
436 smtp_accept_max_per_host allocated resources which were not released
437 when the limit was exceeded. This eventually crashed the daemon. Fix
438 by adding a release action in that path.
440 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
441 expanded; previously using tainted values was rejected. Fix by using
442 dynamically-created buffers.
444 JH/09 Relax restrictions on ACL verify condition needing access to message
445 headers. Previously they were only permitted in data and non-smtp ACLs;
446 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
447 not_blind, header_sender and header_names_ascii verification.
449 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
450 Previously a macro used one argument twice; when called with the
451 argument as an expression having side-effects, incorrect operation
452 resulted. Use an inlineable function.
454 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
455 held open for a verify callout. Previously this wan not accounted for
456 and a corrupt onward SMTP conversation resulted.
458 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
459 Previously they were embedded naively in the constructed address; when
460 needed, strip the quoting and quote the entire local_part.
461 Also make the inbound_srs expansion condition handle quoting.
463 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
464 excluded, not matching the documentation.
466 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
467 was given for the sqlite_dbfile a trap resulted.
469 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
470 "name" argument resulted in a trap. There is no reason to disallow such;
471 this was a coding error.
473 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
474 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
475 spent suspended, ignoring the POSIX definition. Previously we assumed
476 it did and a constant offset from real time could be used as a correction.
477 Change to using the same clock source for the start-of-message and the
478 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
479 exists, just to get a clock slightly more aligned to reality.
481 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
482 RFC says it is optional some validators care. The missing char was not
483 intended but triggered by a line-wrap alignment. Discovery and fix by
484 Guillaume Outters, hacked on by JH.
486 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
487 name being quoted was tainted a trap would be taken. Fix by using
488 dynamically created buffers. The routine could have been called by a
489 rewrite with the "h" flag, by using the "-F" command-line option, or
490 by using a "name=" option on a control=submission ACL modifier.
492 JH/19 SPF: change the Authentication-Results expansion component to give
493 smtp.helo when the sender domain is empty. Previously it gave
496 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
497 not in 127.0.0.0/8 to help in spotting list domains taken over by a
498 domain-parking registrar.
500 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
501 Previously when a whitespace character was specified it was not inserted
502 after removing the newline.
504 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
505 the domain part of the recipient address. This overrides any tls_sni
506 option set, which was previously used.
508 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
511 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
512 is_tainted() had an off-by-one error in the overenthusiastic direction.
513 Find and fix by Gavan. Although NetBSD is not a supported platform for
514 4.94 this bug could affect other platforms.
516 PP/01 Fix default prime selection to be consistent.
517 One path used ike23 still, instead of exim.dev.20160529.3; now both
518 execution flows will use the same DH primes (currently
519 exim.dev.20160529.3).
521 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
522 option in smtp transport, to match the documentation. Previously
523 verification was not being done in some cases where it should have been.
525 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
526 than one server was defined and depending on the platform memory layout
527 details, an internal consistency trap could be hit while walking the list
530 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
531 files. The value is used by the authresults expansion item. Previously
532 if this was used in a router or transport, a crash could result.
534 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
535 have a sufficient incidence of EINTR returns from select that an
536 interaction with other operations done by the main daemon loop exposed
537 a bug in the error-handling. This was benign apart from the log
540 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
541 for consistency with delivered "=>" and failed "**" lines. While we're
542 there, handle PRX and TFO.
544 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
545 applied. This resulted, if any header-line rewrite rules were configured,
546 in a panic-log triggerable by sending a message with a long address in
547 a header. Fix by increasing the arbitrary limit to larger than a single
548 (dewrapped) 5322 header line maximum size.
550 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
551 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
552 RFC 6648 which deprecates X- options in protocols as a general practice.
553 Changeover between the implementations is handled by the mechanisms
556 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
557 a remote which uses callout/hold. Previously the local was queued.
559 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
562 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
563 log line. The asterisk is now consistently appended to the remote IP
564 (and port, if given), and will also be provided on defer and fail log
565 lines. Previously it could be placed on the local IP if that was being
566 logged, and was only provided on delivery lines.
568 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
570 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
571 authenticator client_send option. Previously the next char, after a pair
572 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
573 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
574 documentation. There is still no way to get a leading ^ immediately
575 after a NUL (ie. for the password of a PLAIN method authenticator.
577 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
578 files. For bad sizes read, delete the record and whine to paniclog.
580 JH/38 When logging an AUTH failure, as server, do not include sensitive
581 information. Previously, the credentials would be included if given
582 as part of the AUTH command line and an ACL denied authentication.
584 JH/39 Bug 2691: fix $local_part_data. When the matching list element
585 referred to a file, bad data was returned. This likely also affected
588 JH/40 The gsasl authenticator now supports caching of the salted password
589 generated by the client-side implementation. This required the addition
590 of a new variable: $auth4.
592 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
593 left undeleted; the attempt to re-create it then failed - resulting in
594 the usual "SIGHUP tp have daemon reload configuration" to not work.
595 This affected any platform not supporting "abstract" Unix-domain
596 sockets (i.e. not Linux).
598 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
599 recipients" response to RCPT in a later response, with a 250. The
600 previous coding assumed this would not happen, and under PIPELINING
601 would result in both lost and duplicate recipients for a message.
603 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
604 Previously the weighting was incorrectly applied. Similar fix for socks
605 proxies. Found and fixed by Heiko Schlichting.
607 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
608 not handle sub-lists included using the +namedlist syntax. While
609 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
610 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
613 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
614 Previously the permanent pool was used, so the sore could not be freed.
615 This meant a connection with many messages would use continually-growing
618 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
619 per-pool so as not to waste too much space. Previously a constant size
620 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
621 DOS attacks harder. The cost is wasted memory use in the larger blocks.
623 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
624 same space repeatedly, and a smaller process footprint.
626 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
627 for GnuTLS. Previously the name started "%s" which, while not a bug,
628 looked as if if might be one.
630 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
631 "continued connection") make the $proxy_* variables available. Previously
632 the information was not passed across the exec() call for subsequent
633 transport executions. This also mean that the log lines for the
634 messages can show the proxy information.
636 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
637 receive time. With modern systems the difference is significant.
638 The historical behaviour can be restored by disabling (a new) log_selector
639 "queue_time_exclusive".
641 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
642 content) were specifically tested for. Now, also cover expansions
643 resulting in ACL names and inline ACL content.
645 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
646 making it unusable in complex expressions.
648 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
649 fail with a taint-check on the spoolfile name, and leave the message
652 HS/01 Enforce absolute PID file path name.
654 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
656 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
658 PP/02 Bug 2643: Correct TLS DH constants.
659 A missing NUL termination in our code-generation tool had led to some
660 incorrect Diffie-Hellman constants in the Exim source.
661 Reported by kylon94, code-gen tool fix by Simon Arlott.
663 PP/03 Impose security length checks on various command-line options.
664 Fixes CVE-2020-SPRSS reported by Qualys.
666 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
667 better. Reported by Qualys.
669 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
670 providing a particularly obnoxious sender full name.
673 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
675 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
678 PP/08 Change default for recipients_max from unlimited to 50,000.
680 PP/09 Fix security issue with too many recipients on a message (to remove a
681 known security problem if someone does set recipients_max to unlimited,
682 or if local additions add to the recipient list).
683 Fixes CVE-2020-RCPTL reported by Qualys.
685 PP/10 Fix security issue in SMTP verb option parsing
686 Fixes CVE-2020-EXOPT reported by Qualys.
688 PP/11 Fix security issue in BDAT state confusion.
689 Ensure we reset known-good where we know we need to not be reading BDAT
690 data, as a general case fix, and move the places where we switch to BDAT
691 mode until after various protocol state checks.
692 Fixes CVE-2020-BDATA reported by Qualys.
694 HS/03 Die on "/../" in msglog file names
696 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
697 the Exim runtime user are allowed to create files.
699 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
702 QS/03 When reading the output from interpreted forward files we do not
703 pass the pipe between the parent and the interpreting process to
704 executed child processes (if any).
706 QS/04 Always die if requested from internal logging, even is logging is
709 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
710 the API; compilation noo longer completes with DMARC support included.
711 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
712 been reported on other platforms.
714 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
716 JH/56 Make the majority of info read from config files readonly, for defence-in-
717 depth against exploits. Suggestion by Qualys.
718 Not supported on Solaris 10.
720 JH/57 Fix control=fakreject for a custom message containing tainted data.
721 Previously this resulted in a log complaint, due to a re-expansion present
722 since fakereject was originally introduced.
724 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
725 modifier was given, a loop resulted.
727 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
728 pipelined SMTP command followed the BDAT LAST then it would be
729 incorrectly treated as part of the message body, causing a verification
732 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
733 require looser validation rules than those for 821-level addresses,
734 which only permit IP addresses.
740 JH/01 Avoid costly startup code when not strictly needed. This reduces time
741 for some exim process initialisations. It does mean that the logging
742 of TLS configuration problems is only done for the daemon startup.
744 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
746 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
747 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
750 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
751 DKIM signing is being done. Previously a transport_filter would always
752 disable CHUNKING, falling back to traditional DATA.
754 JH/05 Regard command-line recipients as tainted.
756 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
758 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
759 PAM library frees one of the arguments given to it, despite the
760 documentation. Therefore a plain malloc must be used.
762 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
763 on-stack buffers were used, resulting in a taint trap when DSN information
764 copied from a received message was written into the buffer.
766 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
767 the ordering of its ARC headers. This caused a crash.
769 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
770 a new record was being constructed with information from the peer, a trap
773 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
774 installation would get error messages from DMARC verify, when it hit the
775 nonexistent file indicated by the default. Distros wanting DMARC enabled
776 should both provide the file and set the option.
777 Also enforce no DMARC verification for command-line sourced messages.
779 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
780 could, depending on the platform, hang at the STARTTLS response.
782 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
783 message on a connection. Previously if one message had ARC headers and
784 the following one did not, a crash could result when adding an
785 Authentication-Results: header.
787 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
788 Exim main code and Exim-related utities. The introduction of taint
789 tracking also did many adjustments to string handling. Since then, eximon
790 frequently terminated with an assert failure.
792 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
793 check for 452 responses. This slightly helps the inefficieny of doing
794 a large alias-expansion into a recipient-limited target. The max_rcpt
795 transport option still applies (and at the current default, will override
796 the new feature). The check is done for either cause of synch, and forces
797 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
798 connection. The new facility is not tunable at this time.
800 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
801 library live data was being used, so the results became garbage. Make
802 copies while it is still usable.
804 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
805 on delivery deferred (==) and failed (**) lines (if a delivery was
806 attemtped). Previously it was only on completion (=>) lines.
808 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
809 for the expansion of the server_scram_iter and server_scram_salt options.
811 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
812 are now specifically given a NO_DATA response without hitting the system
813 resolver. The library goes on to do the now-standard TXT lookup.
814 Use of dnsdb lookups is not affected.
816 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
817 only retrieve the errormessage once. Previously two calls to dlerror()
818 were used, and the second one (for mainlog/paniclog) retrieved null
821 JH/20 Taint checking: disallow use of tainted data for
822 - the appendfile transport file and directory options
823 - the pipe transport command
824 - the autoreply transport file, log and once options
825 - file names used by the redirect router (including filter files)
827 - paths used by single-key lookups
828 Previously this was permitted.
830 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
831 adjusted the size of a major service buffer; this failed because the
832 buffer was in use at the time. Change to a compile-time increase in the
833 buffer size, when this authenticator is compiled into exim.
835 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
836 previous fast-mode was untenable in the face of glibs using mmap to
837 support larger malloc requests.
839 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
840 New values supported, if defined on system where compiled:
841 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
842 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
844 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
845 running a limited number of proceses in parallel, a benefit is gained. The
846 amount varies with the platform hardware and load. The use of the option
847 queue_run_in_order means we cannot do this, as ordering becomes
850 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
851 had introduced a string-copy (for ensuring NUL-termination) which was not
852 appropriate for that case, which can include embedded NUL bytes in the
853 block of data. Investigation showed the copy to actually be needless, the
854 data being length-specified.
856 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
857 done during a receiving connection, and both used TLS, global info was
858 used rather than per-connection info for tracking the state of data
859 queued for transmission. This could result in a connection hang.
861 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
862 Previously, when delivering serveral messages down a single connection
863 only the first would provide a SIZE. This was due to the size information
864 not being properly tracked.
866 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
867 TAI (at 37 seconds currently), pretend to be in UTC for time-related
868 expansion and logging. Previously, spurious values such as a future
869 minute could be seen.
871 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
872 it could crash from a null-deref. This could also affect the
873 ${addresses: } operator and ${readsock } item.
875 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
876 message following a mime one, the variable was not reset.
878 JH/30 When an pipelined-connect fails at the first response, assume incorrect
879 cached capability (perhaps the peer reneged?) and immediately retry in
882 JH/31 Fix spurious detection of timeout while writing to transport filter.
884 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
885 an attempt to copy the string was made before checking it.
887 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
888 taint of the lookup key was maintained; we now regard the presence in the
889 filesystem as sufficient validation.
891 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
892 argument is supplied.
894 JH/35 The dsearch lookup now requires that the directory is an absolute path.
895 Previously this was not checked, and nonempty relative paths made an
896 access under Exim's current working directory.
898 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
899 Previously no event was raised.
901 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
902 parameter supplied by the sender MAIL FROM command. Previously it was
903 ignored, and only the check_spool_space option value for the required
906 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
907 the size of the signing public-key. Previously it was instead giving
908 the size of the signature hash.
910 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
911 the default. See the (new) dkim_verify_min_keysizes option.
913 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
914 and an ACL use a lookup for checking either the local_part or domain,
915 stale data could be accessed. Ensure that variable references are
916 dropped between messages.
918 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
919 by the client was not checked as pointing within response data before
920 being used. A malicious client could thus cause an out-of-bounds read and
921 possibly gain authentication. Fix by adding the check.
923 JH/42 Internationalisation: change the default for downconversion in the smtp
924 transport to be "if needed". Previously it was "as previously set" for
925 the message, which usually meant "if needed" for message-submission but
926 "no" for everything else. However, MTAs have been seen using SMTPUTF8
927 even when the envelope addresses did not need it, resulting in forwarding
928 failures to non-supporting MTAs. A downconvert in such cases will be
929 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
930 transport. The change does mean that addresses needing conversion will
931 be converted when previously a delivery failure would occur.
933 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
934 response was received it would be used unchecked in a fail-DSN, violating
935 standards on line-length limits. Truncate if needed.
937 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
938 doesn't work. (Additionally add a new main config option to configure the
945 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
946 side, to decode a TLS 1.3 packet capture.
948 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
949 Previously the default library behaviour applied, sending two, each in
952 JH/03 Debug output for ACL now gives the config file name and line number for
955 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
957 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
959 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
960 buffer overrun for (non-chunking) other transports.
962 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
963 TLS1.3, means that a server rejecting a client certificate is not visible
964 to the client until the first read of encrypted data (typically the
965 response to EHLO). Add detection for that case and treat it as a failed
966 TLS connection attempt, so that the normal retry-in-clear can work (if
967 suitably configured).
969 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
970 and/or domain. Found and fixed by Jason Betts.
972 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
973 configuration). If a CNAME target was not a wellformed name pattern, a
976 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
977 the OS reports them interleaved with other addresses.
979 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
980 used both for input and for a verify callout, both encrypted, SMTP
981 responses being sent by the server could be lost. This resulted in
982 dropped connections and sometimes bounces generated by a peer sending
985 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
986 string. Previously it was possible to cause undefined behaviour in a
987 library routine (usually a crash). Found by "zerons".
989 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
992 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
993 API was removed, so update to use the newer ones.
995 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
996 any timeout set, is taking a long time. Previously we would hang on to a
997 rotated logfile "forever" if the input was arriving with long gaps
998 (a previous attempt to fix addressed lack, for a long time, of initial
1001 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1002 shared (NFS) environment. The length of the tempfile name is now
1003 4 + 16 ("hdr.$message_exim_id") which might break on file
1004 systems which restrict the file name length to lower values.
1005 (It was "hdr.$pid".)
1007 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1008 shared (NFS) environment.
1010 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1011 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1014 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1015 on some platforms for bit 31.
1017 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1018 to changes apparently associated with TLS1.3 handling some of the APIs
1019 previously used were either nonfunctional or inappropriate. Strings
1020 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1021 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1022 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1023 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1024 and the use of specific cipher names in the encrypted= ACL condition.
1026 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1028 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1029 verification result was not updated unless hosts_require_ocsp applied.
1031 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1032 queue_list_requires_admin set to false, non-admin users were denied the
1035 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1036 directory-of-certs mode. Previously they were advertised despite the
1039 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1040 A single TCP connection by a client will now hold a TLS connection open
1041 for multiple message deliveries, by default. Previously the default was to
1044 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1045 default. If built with the facility, DANE will be used. The facility
1046 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1048 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1049 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1050 must be defined and you must still, unless you define DISABLE_TLS, manage
1051 the the include-dir and library-file requirements that go with that
1052 choice. Non-TLS builds are still supported.
1054 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1057 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1058 default. If the platform supports and has the facility enabled, it will
1059 be requested on all coneections.
1061 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1062 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1064 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1066 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1067 success-DSN messages. Previously the From: header was always the default
1068 one for these; the option was ignored.
1070 JH/28 Fix the timeout on smtp response to apply to the whole response.
1071 Previously it was reset for every read, so a teergrubing peer sending
1072 single bytes within the time limit could extend the connection for a
1073 long time. Credit to Qualsys Security Advisory Team for the discovery.
1075 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1076 delivery address, which leaked information of the results of local
1077 forwarding. Change to the original envelope recipient address, per
1080 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1081 requested. Previously not bounce was generated and a log entry of
1082 error ignored was made.
1084 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1086 JH/32 Introduce a general tainting mechanism for values read from the input
1087 channel, and values derived from them. Refuse to expand any tainted
1088 values, to catch one form of exploit.
1090 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1091 was unused and the unexpanded text used for the test. Found and
1092 fixed by Ruben Jenster.
1094 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1095 an attempt to use a TLS library read routine dereffed a nul pointer,
1098 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1099 them in our smtp response.
1101 JH/36 Have the general router option retry_use_local_part default to true when
1102 any of the restrictive preconditions are set (to anything). Previously it
1103 was only for check_local user. The change removes one item of manual
1104 configuration which is required for proper retries when a remote router
1105 handles a subset of addresses for a domain.
1107 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1108 link count into consideration.
1110 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1111 caused the extension of big_buffer, the following lines were ignored.
1113 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1114 accordance with RFC 2308. Previously there was no expiry, so a longlived
1115 receive process (eg. due to ACL delays) versus a short SOA value could
1118 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1120 JH/39 Promote DMARC support to mainline.
1122 JH/40 Bug 2452: Add a References: header to DSNs.
1124 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1125 parameters. The relevant library call is documented as "Deprecated: This
1126 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1127 3.6.0, DH parameters are negotiated following RFC7919."
1129 HS/06 Change the default of dnssec_request_domains to "*"
1131 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1132 carried on and emitted a BDAT command, even when PIPELINING was not
1135 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1136 buffer was used for the filename, resulting in a trap when tainted
1137 arguments (eg. $domain) were used.
1139 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1140 recommended to avoid a possible server-load attack. The feature can be
1141 re-enabled via the openssl_options main cofiguration option.
1143 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1144 for version 4.90 - adding a "more data" boolean to the arguments.
1145 Bumped the ABI version number also, this having been missed previously;
1146 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1147 respect of usage of smtp_printf() by either local_scan code or libraries
1148 accessed via the ${dlfunc } expansion item. Both will need coding
1149 adjustment for any calls to smtp_printf() to match the new function
1150 signature; a FALSE value for the new argument is always safe.
1152 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1153 the file-offset (which the Linux syscall does, and exim expects); this
1154 resulted in an indefinite loop.
1156 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1157 to do ARC verification. The Authentication-Results: header line added
1158 by the configuration then had no ARC item.
1160 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1161 an incorrect mode was used for reading the result, resulting in it being
1168 JH/01 Remove code calling the customisable local_scan function, unless a new
1169 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1171 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1172 non-signal-safe functions being used.
1174 JH/03 Bug 2269: When presented with a received message having a stupidly large
1175 number of DKIM-Signature headers, disable DKIM verification to avoid
1176 a resource-consumption attack. The limit is set at twenty.
1178 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1179 report of oldest_pass in ${authres } in consequence, and separate out
1180 some descriptions of reasons for verification fail.
1182 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1183 files in the spool were present and unlocked. A queue-runner could spot
1184 them, resulting in a duplicate delivery. Fix that by doing the unlock
1185 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1186 add more error-checking on spoolfile handling while that code is being
1189 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1190 No known attacks, no CVE, this is defensive hardening.
1192 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1193 a queue-runner could start a delivery while other operations were ongoing.
1194 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1195 Found and investigated by Tim Stewart. Fix by using the open message data
1196 file handle rather than opening another, and not locally closing it (which
1197 releases a lock) for that case, while creating the temporary .eml format
1198 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1200 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1201 $sender_verify_failure/$recipient_verify_failure to "random".
1203 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1206 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1207 Previously this would segfault.
1209 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1212 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1213 like zero, since the resolver should be doing this for us, But we need one
1214 as a CNAME but no MX presence gets the CNAME returned; we need to check
1215 that doesn't point to an MX to declare it "no MX returned" rather than
1216 "error, loop". A new main option is added so the older capability of
1217 following some limited number of chain links is maintained.
1219 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1221 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1222 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1223 now-deprecated earlier definitions used only specified the range up to TLS
1224 1.2 (in the older-version library docs).
1226 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1228 JH/15 Rework TLS client-side context management. Stop using a global, and
1229 explicitly pass a context around. This enables future use of TLS for
1230 connections to service-daemons (eg. malware scanning) while a client smtp
1231 connection is using TLS; with cutthrough connections this is quite likely.
1233 JH/16 Fix ARC verification to do AS checks in reverse order.
1235 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1237 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1238 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1239 Previously the "utf8" would be re-prepended for every additional message.
1241 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1242 Previously thery were accepted, resulting in issues when attempting to
1243 forward messages to a non-supporting MTA.
1245 PP/02 Let -n work with printing macros too, not just options.
1247 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1248 one parent address was copied, and bogus data was used at delivery-logging
1249 time. Either a crash (after delivery) or bogus log data could result.
1250 Discovery and analysis by Tim Stewart.
1252 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1253 Previously if the string ended mid-character, we did not insert the
1254 promised '?' replacement.
1256 PP/04 Documentation: current string operators work on bytes, not codepoints.
1258 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1259 should pack well giving a smaller memory footprint so better caching and
1260 therefore performance. Group the declarations where this can't be done so
1261 that the byte-sized flag variables are not interspersed among pointer
1262 variables, giving a better chance of good packing by the compiler.
1264 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1265 non-null, to avoid issues with sites running BATV. Previously reports were
1266 sent with an empty envelope sender so looked like bounces.
1268 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1269 The ignore_error flag wasn't being returned from the filter subprocess so
1270 was not set for later routers. Investigation and fix by Matthias Kurz.
1272 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1273 and a msg:complete for the whole, when a message is manually removed using
1274 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1276 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1277 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1278 hope that is portable enough.
1280 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1281 requiring a known-CA anchor certificate; make it now rely entirely on the
1282 TLSA as an anchor. Checking the name on the leaf cert against the name
1283 on the A-record for the host is still done for TA (but not for EE mode).
1285 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1286 included in delivery lines for non-proxied connections, when compiled with
1287 SUPPORT_SOCKS and running with proxy logging enabled.
1289 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1290 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1291 move the existing event to fire before the normal logging of message
1292 failure so that custom logging is bracketed by normal logging.
1294 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1295 msg:fail:internal event. Developement by Matthias Kurz.
1297 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1298 far too small for todays use of crypto signatures stored there. Go all
1299 the way to the max DNS message size of 64kB, even though this might be
1300 overmuch for IOT constrained device use.
1302 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1303 copy a string over itself. The library routine is documented as not
1304 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1306 JH/32 For main options check_spool_space and check_inode_space, where the
1307 platform supports 64b integers, support more than the previous 2^31 kB
1308 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1309 the previous G, M, k.
1311 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1312 $authenticated_fail_id variable on authentication failure. Previously
1315 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1316 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1317 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1318 GNUTLS_SEC_PARAM_MEDIUM.
1320 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1321 side. Previously we would continue as if no SNI had been received.
1323 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1324 "<" character, which should be a list-separator specification, we walked
1325 off past the nul-terimation.
1327 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1328 causes) even when the retry time is not yet met. Previously they were
1329 not, meaning that when (say) an account was over-quota and temp-rejecting,
1330 and multiple senders' messages were queued, only one sender would get
1331 notified on each configured delay_warning cycle.
1333 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1335 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1336 cipher-suites, an error can be left on the stack even for a succeeding
1337 accept; this results in impossible error messages when a later operation
1340 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1341 return error codes indicating retry. Under TLS1.3 this becomes required.
1343 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1344 it only wrote the new authenticators, resulting in a lack of tracking of
1345 peer changes of ESMTP extensions until the next cache flush.
1347 JH/41 Fix the loop reading a message header line to check for integer overflow,
1348 and more-often against header_maxsize. Previously a crafted message could
1349 induce a crash of the recive process; now the message is cleanly rejected.
1351 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1352 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1358 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1359 When redis_servers is set to a list of > 1 element, and the Redis servers
1360 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1361 case of MOVED into a DEFER case instead, thus moving the query onto the
1362 next server in the list. For a cluster of N elements, all N servers must
1363 be defined in redis_servers.
1365 GF/02 Catch and remove uninitialized value warning in exiqsumm
1366 Check for existence of @ARGV before looking at $ARGV[0]
1368 JH/01 Replace the store_release() internal interface with store_newblock(),
1369 which internalises the check required to safely use the old one, plus
1370 the allocate and data copy operations duplicated in both (!) of the
1371 extant use locations.
1373 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1374 modifier. This matches the restriction on the commandline.
1376 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1377 Previously only the last row was returned.
1379 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1380 we assumed that tags in the header were well-formed, and parsed the
1381 element content after inspecting only the first char of the tag.
1382 Assumptions at that stage could crash the receive process on malformed
1385 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1386 While running the DKIM ACL we operate on the Permanent memory pool so that
1387 variables created with "set" persist to the DATA ACL. Also (at any time)
1388 DNS lookups that fail create cache records using the Permanent pool. But
1389 expansions release any allocations made on the current pool - so a dnsdb
1390 lookup expansion done in the DKIM ACL releases the memory used for the
1391 DNS negative-cache, and bad things result. Solution is to switch to the
1392 Main pool for expansions.
1393 While we're in that code, add checks on the DNS cache during store_reset,
1394 active in the testsuite.
1395 Problem spotted, and debugging aided, by Wolfgang Breyha.
1397 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1398 When none of the hosts presented to a transport match an already-open
1399 connection, close it and proceed with the list. Previously we would
1400 queue the message. Spotted by Lena with Yahoo, probably involving
1403 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1404 Previously a spurious "250 OK id=" response was appended to the proper
1407 JH/08 The "support for" informational output now, which built with Content
1408 Scanning support, has a line for the malware scanner interfaces compiled
1409 in. Interface can be individually included or not at build time.
1411 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1412 by the template makefile "src/EDITME". The "STREAM" support for an older
1413 ClamAV interface method is removed.
1415 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1416 rows affected is given instead).
1418 JH/11 The runtime Berkeley DB library version is now additionally output by
1419 "exim -d -bV". Previously only the compile-time version was shown.
1421 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1422 SMTP connection. Previously, when one had more recipients than the
1423 first, an abortive onward connection was made. Move to full support for
1424 multiple onward connections in sequence, handling cutthrough connection
1425 for all multi-message initiating connections.
1427 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1428 routers. Previously, a multi-recipient message would fail to match the
1429 onward-connection opened for the first recipient, and cause its closure.
1431 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1432 a timeout on read on a GnuTLS initiating connection, resulting in the
1433 initiating connection being dropped. This mattered most when the callout
1434 was marked defer_ok. Fix to keep the two timeout-detection methods
1437 JH/15 Relax results from ACL control request to enable cutthrough, in
1438 unsupported situations, from error to silently (except under debug)
1439 ignoring. This covers use with PRDR, frozen messages, queue-only and
1442 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1444 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1445 metadata, resulting in a crash in free().
1447 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1448 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1449 Broken also in d185889f4, with init system revamp.
1451 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1452 Previously we abruptly closed the connection after reading a malware-
1453 found indication; now we go on to read the "scan ok" response line,
1456 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1457 ACL. Previously, a crash would result.
1459 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1460 macro text after a replacement (previously it was only once per line) and
1461 by skipping builtin macros when searching for an uppercase lead character.
1463 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1464 for the build is renamed.
1466 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1467 was allocated for every new TLS startup, meaning one per message. Fix
1468 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1470 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1471 reported the original. Fix to report (as far as possible) the ACL
1472 result replacing the original.
1474 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1475 OpenSSL. Certificate information is loaded for every new TLS startup,
1476 and the resources needed to be freed.
1478 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1480 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1483 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1484 DATA response info to the (existing) per-recipient response info for
1485 the "C=" log element. It can have useful tracking info from the
1486 destination system. Patch from Simon Arlott.
1488 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1489 length value. Previously this would segfault.
1491 HS/02 Support Avast multiline protoocol, this allows passing flags to
1492 newer versions of the scanner.
1494 JH/28 Ensure that variables possibly set during message acceptance are marked
1495 dead before release of memory in the daemon loop. This stops complaints
1496 about them when the debug_store option is enabled. Discovered specifically
1497 for sender_rate_period, but applies to a whole set of variables.
1498 Do the same for the queue-runner and queue-list loops, for variables set
1499 from spool message files. Do the same for the SMTP per-message loop, for
1500 certain variables indirectly set in ACL operations.
1502 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1503 as a multi-recipient message from a mailinglist manager). The coding had
1504 an arbitrary cutoff number of characters while checking for more input;
1505 enforced by writing a NUL into the buffer. This corrupted long / fast
1506 input. The problem was exposed more widely when more pipelineing of SMTP
1507 responses was introduced, and one Exim system was feeding another.
1508 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1509 receiving system, and refused recipients seen by the sending system
1510 (propating to people being dropped from mailing lists).
1511 Discovered and pinpointed by David Carter.
1513 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1514 replaced by the ${authresults } expansion.
1516 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1518 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1519 allows proper process termination in container environments.
1521 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1522 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1524 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1525 and "err_perm", deprecated since 4.83 when the RFC-defined words
1526 "temperror" and "permerror" were introduced.
1528 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1529 transport-filters or DKIM-signing. The restriction was lost in the
1530 consolidation of verify-callout and delivery SMTP handling.
1531 Extend the restriction to also cover ARC-signing.
1533 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1534 in defer=pass mode supply a 450 to the initiator. Previously the message
1537 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1538 tls_require_ciphers is used as before.
1540 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1541 "pass_unscanned". Only tmpfails from the scanner are written to
1542 the paniclog, as they may require admin intervention (permission
1543 denied, license issues). Other scanner errors (like decompression
1544 bombs) do not cause a paniclog entry.
1546 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1547 Previously it was possible to log spurious information in receive log
1550 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1551 triggered odd behaviour from Outlook Express clients.
1553 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1556 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1557 since the IETF WG has not yet settled on that versus the original
1558 "bare" representation.
1560 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1561 Previously the millisecond value corrupted the output.
1562 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1563 corrupted the output.
1569 JH/01 Rework error string handling in TLS interface so that the caller in
1570 more cases is responsible for logging. This permits library-sourced
1571 string to be attached to addresses during delivery, and collapses
1572 pairs of long lines into single ones.
1574 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1575 during configuration. Wildcards are allowed and expanded.
1577 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1578 This permits better logging.
1580 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1581 do not close it down and have the child transport start it up again on
1582 the passed-on TCP connection. Instead, proxy the child (and any
1583 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1584 affected: the continued delivery log lines do not have any DNSSEC, TLS
1585 Certificate or OCSP information. TLS cipher information is still logged.
1587 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1588 identical IP addresses on different listening ports. Will also affect
1591 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1592 add noisy ifdef guards to special-case this sillyness.
1593 Patch from Bernd Kuhls.
1595 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1596 than 255 are no longer allowed.
1598 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1599 Disable the session-cache too, which might reduce our load. Since we
1600 currrectly use a new context for every connection, both as server and
1601 client, there is no benefit for these.
1602 GnuTLS appears to not support tickets server-side by default (we don't
1603 call gnutls_session_ticket_enable_server()) but client side is enabled
1604 by default on recent versions (3.1.3 +) unless the PFS priority string
1607 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1608 <https://reproducible-builds.org/specs/source-date-epoch/>.
1610 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1611 the check for any unsuccessful recipients did not notice the limit, and
1612 erroneously found still-pending ones.
1614 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1615 MSG_MORE. Only in-clear (not on TLS connections).
1617 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1618 transport-filter is involved we can buffer the headers in memory for
1619 creating the signature, and read the spool data file once for the
1620 signature and again for transmission.
1622 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1623 4.77 as the kernel support then wasn't solid, having issues in 64bit
1624 mode. Now, it's been long enough. Add support for FreeBSD also.
1626 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1627 case where the routing stage had gathered several addresses to send to
1628 a host before calling the transport for the first, we previously failed
1629 to close down TLS in the old transport process before passing the TCP
1630 connection to the new process. The new one sent a STARTTLS command
1631 which naturally failed, giving a failed delivery and bloating the retry
1632 database. Investigation and fix prototype from Wolfgang Breyha.
1634 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1635 false-negatives in the check that the sender had not preempted a response
1636 or prompt from Exim (running as a server), due to that code's lack of
1637 awareness of the SMTP input buffering.
1639 PP/04 Add commandline_checks_require_admin option.
1640 Exim drops privileges sanely, various checks such as -be aren't a
1641 security problem, as long as you trust local users with access to their
1642 own account. When invoked by services which pass untrusted data to
1643 Exim, this might be an issue. Set this option in main configuration
1644 AND make fixes to the calling application, such as using `--` to stop
1647 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1648 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1649 responses to those, into a single TLS record each way (this usually means
1650 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1653 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1654 affects you only if you're dancing at the edge of the param size limits.
1655 If you are, and this message makes sense to you, then: raise the
1656 configured limit or use OpenSSL 1.1. Nothing we can do for older
1659 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1660 cmdline element to get the documented default one. Previously it was
1663 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1664 are unsuable for non-message-specific reasons - eg. connection timeout,
1665 banner-time rejection.
1667 JH/16 Fix logging of delivery remote port, when specified by router, under
1670 PP/06 Repair manualroute's ability to take options in any order, even if one
1671 is the name of a transport.
1674 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1676 JH/17 Change the list-building routines interface to use the expanding-string
1677 triplet model, for better allocation and copying behaviour.
1679 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1680 Previously it was constructed the first time a possibly-matching string
1681 was met in the configuration file input during startup; now it is done
1684 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1685 compatible one, to avoid the (poorly documented) possibility of a config
1686 file in the working directory redirecting the DB files, possibly correpting
1687 some existing file. CVE-2017-10140 assigned for BDB.
1689 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1690 cache-hot. Previously, although the result was properly cached, the
1691 initial verify call returned a defer.
1693 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1694 the main verify for receipient in uncached-mode.
1696 JH/22 Retire historical build files to an "unsupported" subdir. These are
1697 defined as "ones for which we have no current evidence of testing".
1699 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1700 if present. Previously it was ignored.
1702 JH/24 Start using specified-initialisers in C structure init coding. This is
1703 a C99 feature (it's 2017, so now considered safe).
1705 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1706 if was a fixed-sized field and bitmask ops via macros; it is now more
1709 PP/07 GitHub PR 56: Apply MariaDB build fix.
1710 Patch provided by Jaroslav Škarvada.
1712 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1713 during Coverity cleanups [4.87 JH/47]
1714 Diagnosis and fix provided by Michael Fischer v. Mollard.
1716 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1717 the right size to place the terminating semicolon on its own folded
1718 line, the header hash was calculated to an incorrect value thanks to
1719 the (relaxed) space the fold became.
1721 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1722 and confused the parent.
1724 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1725 which could crash as a result. This could lead to undeliverable messages.
1727 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1730 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1731 matters on fast-turnover and PID-randomising systems, which were getting
1732 out-of-order delivery.
1734 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1735 a possibly-overlapping copy. The symptom was that "Remote host closed
1736 connection in response to HELO" was logged instead of the actual 4xx
1739 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1740 Previously only that bufferd was discarded, resulting in SYMTP command
1743 JH/32 DKIM: when a message has multiple signatures matching an identity given
1744 in dkim_verify_signers, run the dkim acl once for each. Previously only
1745 one run was done. Bug 2189.
1747 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1748 "panic the current process" to "deliberately defer". The panic log is
1749 still written with the problem list name; the mail and reject logs now
1750 get a temp-reject line for the message that was being handled, saying
1751 something like "domains check lookup or other defer". The SMTP 451
1752 message is still "Temporary local problem".
1754 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1755 A crafted sequence of BDAT commands could result in in-use memory beeing
1756 freed. CVE-2017-16943.
1758 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1759 from SMTP input. Previously it was always done; now only done for DATA
1760 and not BDAT commands. CVE-2017-16944.
1762 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1763 to the message (such as an overlong header line). Previously this was
1764 not done and we did not exit BDAT mode. Followon from the previous item
1765 though a different problem.
1771 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1772 than -2003 did; needs libidn2 in addition to libidn.
1774 JH/02 The path option on a pipe transport is now expanded before use.
1776 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1777 Patch provided by "Björn", documentation fix added too.
1779 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1780 missing a wire-to-host endian conversion.
1782 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1783 close after a BDAT command line could be taken as a following command,
1784 giving a synch failure. Fix by only checking for synch immediately
1785 before acknowledging the chunk.
1787 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1788 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1789 macro. Patches provided by Josh Soref.
1791 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1792 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1793 by IANA as a service extension. However, John Klensin suggests that we
1796 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1797 the dkim code may be unix-mode line endings rather than smtp wire-format
1798 CRLF, so prepend a CR to any bare LF.
1800 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1801 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1803 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1804 the first were themselves being wrongly included in the feed into dkim
1805 processing; with most chunk sizes in use this resulted in an incorrect
1806 body hash calculated value.
1808 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1809 DKIM signature block, for verification. Although advised against by
1810 standards it is specifically not ruled illegal.
1812 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1814 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1815 missing a body hash (the bh= tag).
1817 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1818 It seems that HAProxy sends the Proxy Protocol information in clear and
1819 only then does a TLS startup, so do the same.
1821 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1822 TCP connections (such as for Spamd) unless the daemon successfully set
1823 Fast Open mode on its listening sockets. This fixes breakage seen on
1824 too-old kernels or those not configured for Fast Open, at the cost of
1825 requiring both directions being enabled for TFO, and TFO never being used
1826 by non-daemon-related Exim processes.
1828 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1829 endings, at least on the first header line. Try to canonify any that get
1830 past that check, despite the cost.
1832 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1833 now limited to an arbitrary five deep, while parsing addresses with the
1834 strip_excess_angle_brackets option enabled.
1836 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1837 instead leave the unprompted TLS handshake in socket buffer for the
1838 TLS library to consume.
1840 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1842 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1844 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1845 region in one operation, for speed, and this leaves assigned pointers
1846 dangling. Add checks run only under the testsuite which checks all
1847 variables at a store-reset and panics on a dangling pointer; add code
1848 explicitly nulling out all the variables discovered. Fixes one known
1849 bug: a transport crash, where a dangling pointer for $sending_ip_address
1850 originally assigned in a verify callout, is re-used.
1852 PP/06 Drop '.' from @INC in various Perl scripts.
1854 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1856 PP/08 Reduce a number of compilation warnings under clang; building with
1857 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1858 should be warning-free.
1860 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1862 HS/01 Fix portability problems introduced by PP/08 for platforms where
1863 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1865 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1866 chunk. This allows us to accept broken chunked messages. We need a more
1867 general solution here.
1869 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1870 already-broken messages in the queue.
1872 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1874 JH/19 Fix reference counting bug in routing-generated-address tracking.
1880 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1881 supports it and a size is available (ie. the sending peer gave us one).
1883 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1884 years of being deprecated). The replacements are the ACLs
1885 acl_smtp_mime and acl_not_smtp_mime.
1887 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1888 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1889 or even an in-clear connection were permitted. Now, if the host lookup
1890 was dnssec and dane was requested then the host is only used if the
1891 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1892 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1893 if one fails this test.
1894 This means that a poorly-configured remote DNS will make it incommunicado;
1895 but it protects against a DNS-interception attack on it.
1897 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1898 non-noisy when a race steals the message being considered.
1900 JH/05 If main configuration option tls_certificate is unset, generate a
1901 self-signed certificate for inbound TLS connections.
1903 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1904 in rewrites and routers.
1906 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1907 and logged a warning sing 4.83; now they are a configuration file error.
1909 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1910 (lacking @domain). Apply the same qualification processing as RCPT.
1912 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1914 JH/10 Support ${sha256:} applied to a string (as well as the previous
1917 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1918 a cutthrough deliver is pending, as we always want to make a connection.
1919 This also avoids re-routing the message when later placing the cutthrough
1920 connection after a verify cache hit.
1921 Do not update it with the verify result either.
1923 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1924 when routing results in more than one destination address.
1926 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1927 signing (which inhibits the cutthrough capability). Previously only
1928 the presence of an option was tested; now an expansion evaluating as
1929 empty is permissible (obviously it should depend only on data available
1930 when the cutthrough connection is made).
1932 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1933 the relevant preceding SMTP command did not note the pipelining mode.
1935 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1936 Previously they were not counted.
1938 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1939 as one having no matching records. Previously we deferred the message
1940 that needed the lookup.
1942 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1943 distinguished as "(=".
1945 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1946 for missing MX records. Previously it only worked for missing A records.
1948 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1950 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1951 after the data-go-ahead and data-ack. Patch from Jason Betts.
1953 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1954 even for a "none" policy. Patch from Tony Meyer.
1956 JH/22 Fix continued use of a connection for further deliveries. If a port was
1957 specified by a router, it must also match for the delivery to be
1960 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1961 When one of the recipients of a message was unsuitable for the connection
1962 (has no matching addresses), we lost track of needing to mark it
1963 deferred. As a result mail would be lost.
1965 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1967 JH/25 Decoding ACL controls is now done using a binary search; the source code
1968 takes up less space and should be simpler to maintain. Merge the ACL
1969 condition decode tables also, with similar effect.
1971 JH/26 Fix problem with one_time used on a redirect router which returned the
1972 parent address unchanged. A retry would see the parent address marked as
1973 delivered, so not attempt the (identical) child. As a result mail would
1976 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1977 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1978 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1981 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1982 Main config options check_{log,spool}_{inodes,space} are now
1983 100 inodes, 10MB unless set otherwise in the configuration.
1985 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1986 Previously it only applied to the main-section connection policy
1989 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1991 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1992 by me. Added RFC7919 DH primes as an alternative.
1994 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1995 are not in the system include path.
1997 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1998 GnuTLS, when a session startup failed (eg because the client disconnected)
1999 Exim did stdio operations after fclose. This was exposed by a recent
2000 change which nulled out the file handle after the fclose.
2002 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2003 signed directly by the cert-signing cert, rather than an intermediate
2004 OCSP-signing cert. This is the model used by LetsEncrypt.
2006 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2008 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2009 an incoming connection.
2011 HS/02 Bug 1802: Do not half-close the connection after sending a request
2014 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2015 fallback to "prime256v1".
2017 JH/34 SECURITY: Use proper copy of DATA command in error message.
2018 Could leak key material. Remotely exploitable. CVE-2016-9963.
2024 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2025 and 3.4.4 - once the server is enabled to respond to an OCSP request
2026 it does even when not requested, resulting in a stapling non-aware
2027 client dropping the TLS connection.
2029 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2030 support variable-length bit vectors. No functional change.
2032 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2033 The I= interface field on outgoing lines is now after the H= remote
2034 host field, same as incoming lines. There is a separate
2035 outgoing_interface log selector which allows you to disable the
2038 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2039 If not running log_selector +smtp_connection the mainlog would be held
2040 open indefinitely after a "too many connections" event, including to a
2041 deleted file after a log rotate. Leave the per net connection logging
2042 leaving it open for efficiency as that will be quickly detected by the
2043 check on the next write.
2045 HS/01 Bug 1671: Fix post transport crash.
2046 Processing the wait-<transport> messages could crash the delivery
2047 process if the message IDs didn't exist for some reason. When
2048 using 'split_spool_directory=yes' the construction of the spool
2049 file name failed already, exposing the same netto behaviour.
2051 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2052 mime_regex ACL conditions.
2054 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2055 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2056 to HELO, local diagnostic string.
2058 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2059 log line to debug. Even when configured with a tls authenticator many
2060 client connections are expected to not authenticate in this way, so
2061 an authenticate fail is not an error.
2063 HS/02 Add the Exim version string to the process info. This way exiwhat
2064 gives some more detail about the running daemon.
2066 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2067 matter for fast-change records such as DNSBLs.
2069 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2070 retry record, even if constant. There may be multiple transports with
2071 different interface settings and the retry behaviour needs to be kept
2074 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2076 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2078 JH/10 Bug 840: fix log_defer_output option of pipe transport
2080 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2081 hosts_randomize. This matters a lot when combined with mua_wrapper.
2083 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2084 ${quote_pgsql:<string>} operator.
2086 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2088 JH/14 Change header file name for freeradius-client. Relevant if compiling
2089 with Radius support; from the Gentoo tree and checked under Fedora.
2091 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2093 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2094 option transport_filter, meaning no filtering.
2096 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2098 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2099 defaults to "*" (all hosts). The variable is now available when not built
2100 with TLS, default unset, mainly to enable keeping the testsuite sane.
2101 If a server certificate is not supplied (via tls_certificate) an error is
2102 logged, and clients will find TLS connections fail on startup. Presumably
2103 they will retry in-clear.
2104 Packagers of Exim are strongly encouraged to create a server certificate
2105 at installation time.
2107 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2108 with the $config_file variable.
2110 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2111 in transport context, after the attempt, and per-recipient. The latter type
2112 is per host attempted. The event data is the error message, and the errno
2113 information encodes the lookup type (A vs. MX) used for the (first) host,
2114 and the trailing two digits of the smtp 4xx response.
2116 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2117 to write to mainlog (or rejectlog, paniclog) in the window between file
2118 creation and permissions/ownership being changed. Particularly affects
2119 installations where exicyclog is run as root, rather than exim user;
2120 result is that the running daemon panics and dies.
2122 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2124 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2125 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2126 "pri" and "weight". Note that the previous implicit priority given by the
2127 list order is no longer honoured.
2129 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2130 for DKIM processing.
2132 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2133 by defining SUPPORT_SOCKS.
2135 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2136 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2137 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2138 variables are renamed to proxy_{local,external}_{address,port}.
2140 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2141 for a build by defining SUPPORT_I18N
2143 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2144 of the query string, and make ${quote_redis:} do that quoting.
2146 JH/29 Move Events support from Experimental to mainline, enabled by default
2147 and removable for a build by defining DISABLE_EVENT.
2149 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2151 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2152 cached by the daemon.
2154 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2155 by defining LOOKUP_REDIS. The libhiredis library is required.
2157 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2158 keys are given for lookup.
2160 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2161 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2162 only supported when built with TLS support. The PolarSSL SHA routines
2163 are still used when the TLS library is too old for convenient support.
2165 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2166 openssl_options), for security. OpenSSL forces this from version 1.1.0
2167 server-side so match that on older versions.
2169 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2170 allocation for $value could be released as the expansion processing
2171 concluded, but leaving the global pointer active for it.
2173 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2174 and to use the domains and local_parts ACL conditions.
2176 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2177 incorrectly not doubled on cutthrough transmission, hence seen as a
2178 body-termination at the receiving system - resulting in truncated mails.
2179 Commonly the sender saw a TCP-level error, and retransmitted the message
2180 via the normal store-and-forward channel. This could result in duplicates
2181 received - but deduplicating mailstores were liable to retain only the
2182 initial truncated version.
2184 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2186 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2188 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2189 we're in there, support oversigning also; bug 1309.
2191 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2193 HS/04 Add support for keep_environment and add_environment options.
2195 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2196 either intentional arithmetic overflow during PRNG, or testing config-
2199 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2200 delivery resulted in actual delivery. Cancel cutthrough before DATA
2203 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2204 rejects a recipient: pass the reject to the originator.
2206 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2207 Many were false-positives and ignorable, but it's worth fixing the
2210 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2211 for the new environment-manipulation done at startup. Move the routines
2212 from being local to tls.c to being global via the os.c file.
2214 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2215 an extract embedded as result-arg for a map, the first arg for extract
2216 is unavailable so we cannot tell if this is a numbered or keyed
2217 extraction. Accept either.
2223 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2226 JH/02 The smtp transport option "multi_domain" is now expanded.
2228 JH/03 The smtp transport now requests PRDR by default, if the server offers
2231 JH/04 Certificate name checking on server certificates, when exim is a client,
2232 is now done by default. The transport option tls_verify_cert_hostnames
2233 can be used to disable this per-host. The build option
2234 EXPERIMENTAL_CERTNAMES is withdrawn.
2236 JH/05 The value of the tls_verify_certificates smtp transport and main options
2237 default to the word "system" to access the system default CA bundle.
2238 For GnuTLS, only version 3.0.20 or later.
2240 JH/06 Verification of the server certificate for a TLS connection is now tried
2241 (but not required) by default. The verification status is now logged by
2242 default, for both outbound TLS and client-certificate supplying inbound
2245 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2248 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2249 Status Notification (bounce) messages are now MIME format per RFC 3464.
2250 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2251 under the control of the dsn_advertise_hosts option, and routers may
2252 have a dsn_lasthop option.
2254 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2255 default, modifiable by a malware= option. The list separator for
2256 the options can now be changed in the usual way. Bug 68.
2258 JH/10 The smtp_receive_timeout main option is now expanded before use.
2260 JH/11 The incoming_interface log option now also enables logging of the
2261 local interface on delivery outgoing connections.
2263 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2264 if the interface and destination host and port all match.
2266 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2269 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2270 Patch from Andrew Lewis.
2272 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2273 now supports optional time-restrictions, weighting, and priority
2274 modifiers per server. Patch originally by <rommer@active.by>.
2276 JH/16 The spamd_address main option now supports a mixed list of local
2277 and remote servers. Remote servers can be IPv6 addresses, and
2278 specify a port-range.
2280 JH/17 Bug 68: The spamd_address main option now supports an optional
2281 timeout value per server.
2283 JH/18 Bug 1581: Router and transport options headers_add/remove can
2284 now have the list separator specified.
2286 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2289 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2292 JH/21 Support for the A6 type of dns record is withdrawn.
2294 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2295 rather than the verbs used.
2297 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2298 from 255 to 1024 chars.
2300 JH/24 Verification callouts now attempt to use TLS by default.
2302 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2303 are generic router options now. The defaults didn't change.
2305 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2306 Original patch from Alexander Shikoff, worked over by JH.
2308 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2309 files if ZCAT_COMMAND is not executable.
2311 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2313 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2315 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2316 Normally benign, it bites when the pair was led to by a CNAME;
2317 modern usage is to not canonicalize the domain to a CNAME target
2318 (and we were inconsistent anyway for A-only vs AAAA+A).
2320 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2322 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2323 when evaluating $sender_host_dnssec.
2325 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2326 $sender_helo_dnssec variable.
2328 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2330 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2332 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2334 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2335 documented as working, but never had. Support all but $spam_report.
2337 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2338 added for tls authenticator.
2340 HS/03 Add perl_taintmode main config option
2346 TL/01 When running the test suite, the README says that variables such as
2347 no_msglog_check are global and can be placed anywhere in a specific
2348 test's script, however it was observed that placement needed to be near
2349 the beginning for it to behave that way. Changed the runtest perl
2350 script to read through the entire script once to detect and set these
2351 variables, reset to the beginning of the script, and then run through
2352 the script parsing/test process like normal.
2354 TL/02 The BSD's have an arc4random API. One of the functions to induce
2355 adding randomness was arc4random_stir(), but it has been removed in
2356 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2357 function when detected.
2359 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2360 cause callback expansion.
2362 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2363 syntax errors in an expansion can be treated as a string instead of
2364 logging or causing an error, due to the internal use of bool_lax
2365 instead of bool when processing it.
2367 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2368 server certificates when making smtp deliveries.
2370 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2372 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2374 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2376 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2377 Merged patch from Sebastian Wiedenroth.
2379 JH/05 Fix results-pipe from transport process. Several recipients, combined
2380 with certificate use, exposed issues where response data items split
2381 over buffer boundaries were not parsed properly. This eventually
2382 resulted in duplicates being sent. This issue only became common enough
2383 to notice due to the introduction of connection certificate information,
2384 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2386 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2387 size buffer was used, resulting in syntax errors when an expansion
2390 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2391 version 3.3.6 or later.
2393 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2394 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2395 both become "event_action", the variables become $event_name, $event_data
2396 and $event_defer_errno. There is a new variable $verify_mode, usable in
2397 routers, transports and related events. The tls:cert event is now also
2398 raised for inbound connections, if the main configuration event_action
2401 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2402 early OCSP support, but no stapling (appears to be less than 1.0.0).
2404 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2405 server certificate names available under the smtp transport option
2406 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2409 JH/10 Time-related extraction expansions from certificates now use the main
2410 option "timezone" setting for output formatting, and are consistent
2411 between OpenSSL and GnuTLS compilations. Bug 1541.
2413 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2414 encoded parameter in the incoming message. Bug 1558.
2416 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2417 include certificate info, eximon was claiming there were spoolfile
2420 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2422 JH/14 Log delivery-related information more consistently, using the sequence
2423 "H=<name> [<ip>]" wherever possible.
2425 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2426 are problematic for Debian distribution, omit them from the release
2429 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2431 JH/16 Fix string representation of time values on 64bit time_t architectures.
2434 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2435 output list separator was used.
2440 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2441 checkers that were complaining about end of non-void function with no
2444 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2445 This was a regression introduced in 4.83 by another bugfix.
2447 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2449 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2450 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2456 TF/01 Correctly close the server side of TLS when forking for delivery.
2458 When a message was received over SMTP with TLS, Exim failed to clear up
2459 the incoming connection properly after forking off the child process to
2460 deliver the message. In some situations the subsequent outgoing
2461 delivery connection happened to have the same fd number as the incoming
2462 connection previously had. Exim would try to use TLS and fail, logging
2463 a "Bad file descriptor" error.
2465 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2466 utilities have not been installed.
2468 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2469 temporary space as the ACL may create new global variables.
2471 TL/01 LDAP support uses per connection or global context settings, depending
2472 upon the detected version of the libraries at build time.
2474 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2475 to extract and use the src ip:port in logging and expansions as if it
2476 were a direct connection from the outside internet. PPv2 support was
2477 updated based on HAProxy spec change in May 2014.
2479 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2481 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2482 Properly escape header and check for NULL return.
2484 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2487 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2489 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2490 characters in header names, implemented as a verify condition.
2491 Contributed by Michael Fischer v. Mollard.
2493 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2494 results permerror and temperror. Previous values are deprecated but
2495 still accepted. In a future release, err_perm and err_temp will be
2496 completely removed, which will be a backward incompatibility if the
2497 ACL tests for either of these two old results. Patch contributed by
2498 user bes-internal on the mailing list.
2500 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2502 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2503 selectors, in both main and reject logs.
2505 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2508 JH/07 Add malware type "sock" for talking to simple daemon.
2510 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2512 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2513 routers/transports under cutthrough routing.
2515 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2516 numbers. Touch up "bool" conditional to keep the same definition.
2518 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2520 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2522 JH/12 Expand items in router/transport headers_add or headers_remove lists
2523 individually rather than the list as a whole. Bug 1452.
2525 Required for reasonable handling of multiple headers_ options when
2526 they may be empty; requires that headers_remove items with embedded
2527 colons must have them doubled (or the list-separator changed).
2529 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2530 view the policy declared in the DMARC record. Currently, $dmarc_status
2531 is a combined value of both the record presence and the result of the
2534 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2536 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2537 dnslookup router and the smtp transport (applying to the forward
2540 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2541 of ldap servers used for a specific lookup. Patch provided by Heiko
2544 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2545 New variable $lookup_dnssec_authenticated for observability.
2547 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2548 Patch submitted by Lars Timman.
2550 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2552 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2553 Requires trusted mode and valid format message id, aborts otherwise.
2554 Patch contributed by Heiko Schlichting.
2556 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2557 certextract with support for various fields. Bug 1358.
2559 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2560 is requested by default, modifiable by smtp transport option
2563 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2564 operate on certificate variables to give certificate fingerprints
2565 Also new ${sha256:cert_variable}.
2567 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2569 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2570 Christian Aistleitner.
2572 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2574 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2575 file. Patch from Wolfgang Breyha.
2577 JH/25 Expand the coverage of the delivery $host and $host_address to
2578 client authenticators run in verify callout. Bug 1476.
2580 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2581 align with daemon_smtp_ports. Bug 72.
2583 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2584 support and error reporting did not work properly.
2586 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2587 and is readable. Patch from Andrew Colin Kissa.
2589 TL/14 Enhance documentation of ${run expansion and how it parses the
2590 commandline after expansion, particularly in the case when an
2591 unquoted variable expansion results in an empty value.
2593 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2595 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2596 by renaming away from it. Addresses GH issue 15, reported by
2599 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2601 TL/15 SECURITY: prevent double expansion in math comparison functions
2602 (can expand unsanitized data). Not remotely exploitable.
2609 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2611 PP/02 Make -n do something, by making it not do something.
2612 When combined with -bP, the name of an option is not output.
2614 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2617 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2618 $sender_host_name and config options to manage this, and basic check
2621 PP/05 DSCP support for outbound connections and control modifier for inbound.
2623 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2624 (Only plugin which currently uses this is kerberos4, which nobody should
2625 be using, but we should make it available and other future plugins might
2626 conceivably use it, even though it would break NAT; stuff *should* be
2627 using channel bindings instead).
2629 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2630 name; added for Sendmail compatibility; requires admin caller.
2631 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2632 just ignore it); requires trusted caller.
2633 Also parse but ignore: -Ac -Am -X<logfile>
2636 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2638 TL/02 Add +smtp_confirmation as a default logging option.
2640 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2641 Patch by Magnus Holmgren from 2007-02-20.
2643 TL/04 Bugzilla 1281 - Spec typo.
2644 Bugzilla 1283 - Spec typo.
2645 Bugzilla 1290 - Spec grammar fixes.
2647 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2649 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2651 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2652 mailing list by by Dmitry Isaikin.
2654 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2656 JH/02 Support "G" suffix to numbers in ${if comparisons.
2658 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2660 NM/01 Bugzilla 1197 - Spec typo
2661 Bugzilla 1196 - Spec examples corrections
2663 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2665 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2666 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2669 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2670 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2672 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2673 "acl {{name}{arg}...}", and optional args on acl condition
2676 JH/05 Permit multiple router/transport headers_add/remove lines.
2678 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2680 JH/07 Avoid using a waiting database for a single-message-only transport.
2681 Performance patch from Paul Fisher. Bugzilla 1262.
2683 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2686 JH/09 Add $headers_added variable, with content from use of ACL modifier
2687 add_header (but not yet added to the message). Bugzilla 199.
2689 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2690 Pulled from Bugzilla 817 by Wolfgang Breyha.
2692 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2694 (nb: this is the same fix as in Exim 4.80.1)
2696 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2699 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2702 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2704 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2705 advertise SMTP AUTH mechanism to us, instead of a generic
2706 protocol violation error. Also, make Exim more robust to bad
2707 data from the Dovecot auth socket.
2709 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2711 When a queue runner is handling a message, Exim first routes the
2712 recipient addresses, during which it prunes them based on the retry
2713 hints database. After that it attempts to deliver the message to
2714 any remaining recipients. It then updates the hints database using
2717 So if a recipient address works intermittently, it can get repeatedly
2718 deferred at routing time. The retry hints record remains fresh so the
2719 address never reaches the final cutoff time.
2721 This is a fairly common occurrence when a user is bumping up against
2722 their storage quota. Exim had some logic in its local delivery code
2723 to deal with this. However it did not apply to per-recipient defers
2724 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2726 This change adds a proper retry rule check during routing so that the
2727 final cutoff time is checked against the message's age. We only do
2728 this check if there is an address retry record and there is not a
2729 domain retry record; this implies that previous attempts to handle
2730 the address had the retry_use_local_parts option turned on. We use
2731 this as an approximation for the destination being like a local
2732 delivery, as in LMTP.
2734 I suspect this new check makes the old local delivery cutoff check
2735 redundant, but I have not verified this so I left the code in place.
2737 TF/02 Correct gecos expansion when From: is a prefix of the username.
2739 Test 0254 submits a message to Exim with the header
2743 When I ran the test suite under the user fanf2, Exim expanded
2744 the header to contain my full name, whereas it should have added
2745 a Resent-Sender: header. It erroneously treats any prefix of the
2746 username as equal to the username.
2748 This change corrects that bug.
2750 GF/01 DCC debug and logging tidyup
2751 Error conditions log to paniclog rather than rejectlog.
2752 Debug lines prefixed by "DCC: " to remove any ambiguity.
2754 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2756 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2757 Bug spotted by Jeremy Harris; was flawed since initial commit.
2758 Would have resulted in OCSP responses post-SNI triggering an Exim
2759 NULL dereference and crash.
2761 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2763 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2764 Bug detection, analysis and fix by Samuel Thibault.
2765 Bugzilla 1331, Debian bug #698092.
2767 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2769 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2770 Server implementation by Todd Lyons, client by JH.
2771 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2772 config variable "prdr_enable" controls whether the server
2773 advertises the facility. If the client requests PRDR a new
2774 acl_data_smtp_prdr ACL is called once for each recipient, after
2775 the body content is received and before the acl_smtp_data ACL.
2776 The client is controlled by both of: a hosts_try_prdr option
2777 on the smtp transport, and the server advertisement.
2778 Default client logging of deliveries and rejections involving
2779 PRDR are flagged with the string "PRDR".
2781 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2782 fclose(). Diagnosis by Todd Lyons.
2784 PP/17 Update configure.default to handle IPv6 localhost better.
2785 Patch by Alain Williams (plus minor tweaks).
2788 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2789 This is now consistent with GnuTLS, and is now documented: the
2790 previous undocumented portable approach to treating the option as
2791 unset was to force an expansion failure. That still works, and
2792 an empty string is now equivalent.
2794 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2795 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2796 not performing validation itself.
2798 PP/20 Added force_command boolean option to pipe transport.
2799 Patch from Nick Koston, of cPanel Inc.
2801 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2804 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2806 PP/21 Fix eximon continuous updating with timestamped log-files.
2807 Broken in a format-string cleanup in 4.80, missed when I repaired the
2808 other false fix of the same issue.
2809 Report and fix from Heiko Schlichting.
2812 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2813 Report from Prashanth Katuri.
2815 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2816 It's SecureTransport, so affects any MacOS clients which use the
2817 system-integrated TLS libraries, including email clients.
2819 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2820 using a MIME ACL for non-SMTP local injection.
2821 Report and assistance in diagnosis by Warren Baker.
2823 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2825 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2827 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2828 last id that failed so it may be referenced in subsequent ACL's.
2830 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2833 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2834 ldap library initialization, allowing self-signed CA's to be
2835 used. Also properly sets require_cert option later in code by
2836 using NULL (global ldap config) instead of ldap handle (per
2837 session). Bug diagnosis and testing by alxgomz.
2839 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2840 the src/util/ subdirectory.
2842 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2843 renamed to Transport Post Delivery Action by Jeremy Harris, as
2846 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2847 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2848 redis_servers = needs to be configured which will be used by the redis
2849 lookup. Patch from Warren Baker, of The Packet Hub.
2851 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2853 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2854 hostname or reverse DNS when processing a host list. Used suggestions
2855 from multiple comments on this bug.
2857 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2859 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2860 Missed a few lines, added it to make the runtest require no keyboard
2863 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2864 contains upper case chars. Make router use caseful_local_part.
2866 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2867 support when GnuTLS has been built with p11-kit.
2873 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2875 This, or similar/improved, will also be change PP/11 of 4.82.
2881 PP/01 Handle short writes when writing local log-files.
2882 In practice, only affects FreeBSD (8 onwards).
2883 Bugzilla 1053, with thanks to Dmitry Isaikin.
2885 NM/01 Bugzilla 949 - Documentation tweak
2887 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2890 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2892 PP/02 Implemented gsasl authenticator.
2894 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2896 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2897 `pkg-config foo` for cflags/libs.
2899 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2900 with rest of GSASL and with heimdal_gssapi.
2902 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2903 `pkg-config foo` for cflags/libs for the TLS implementation.
2905 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2906 properties get this fed in as external SSF. A number of robustness
2907 and debugging improvements to the cyrus_sasl authenticator.
2909 PP/08 cyrus_sasl server now expands the server_realm option.
2911 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2912 Patch by Jeremy Harris.
2914 PP/10 Added dbmjz lookup type.
2916 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2918 PP/12 MAIL args handles TAB as well as SP, for better interop with
2919 non-compliant senders.
2920 Analysis and variant patch by Todd Lyons.
2922 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2923 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2924 Patch from Dirk Mueller <dmueller@suse.com>
2926 PP/13 tls_peerdn now print-escaped for spool files.
2927 Observed some $tls_peerdn in wild which contained \n, which resulted
2928 in spool file corruption.
2930 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2931 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2932 or write after TLS renegotiation, which otherwise led to messages
2935 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2936 as a tracking header (ie: a signed header comes before the signature).
2937 Patch from Wolfgang Breyha.
2939 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2940 comma-sep list; embedded commas doubled.
2942 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2944 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2946 Report and patch from Dmitry Banschikov.
2948 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2949 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2950 failures. We appear to now support TLS1.1+ with Exim.
2952 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2953 lets Exim select keys and certificates based upon TLS SNI from client.
2954 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2955 before an outbound SMTP session. New log_selector, +tls_sni.
2957 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2958 NULL dereference. Report and patch from Alun Jones.
2960 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2961 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2962 Not seeing resolver debug output on NetBSD, but suspect this is a
2963 resolver implementation change.
2965 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2966 Left warnings. Added "eximon gdb" invocation mode.
2968 PP/21 Defaulting "accept_8bitmime" to true, not false.
2970 PP/22 Added -bw for inetd wait mode support.
2972 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2973 locate the relevant includes and libraries. Made this the default.
2975 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2976 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2978 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2979 This may cause build issues on older platforms.
2981 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2982 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2983 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2984 Added SNI support via GnuTLS too.
2985 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2987 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2989 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2990 Applied second patch from Janne, implementing suggestion to default
2991 multiple-strings-in-record handling to match SPF spec.
2993 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2995 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2996 read-only, out of scope).
2997 Patch from Wolfgang Breyha, report from Stuart Northfield.
2999 PP/29 Fix three issues highlighted by clang analyser static analysis.
3000 Only crash-plausible issue would require the Cambridge-specific
3001 iplookup router and a misconfiguration.
3002 Report from Marcin Mirosław.
3004 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3006 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3007 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3008 As part of this, removing so much warning spew let me fix some minor
3009 real issues in debug logging.
3011 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3012 assignment on my part. Fixed.
3014 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3015 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3016 Janne Snabb (who went above and beyond: thank you).
3018 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3019 string otherwise requires a connection and a bunch more work and it's
3020 relatively easy to get wrong. Should also expose TLS library linkage
3023 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3024 64-bit ${eval} (JH/03).
3026 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3027 GNU libc to support some of the 64-bit stuff, should not lead to
3028 conflicts. Defined before os.h is pulled in, so if a given platform
3029 needs to override this, it can.
3031 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3032 protection layer was required, which is not implemented.
3033 Bugzilla 1254, patch from Wolfgang Breyha.
3035 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3036 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3037 tls_dhparam take prime identifiers. Also unbreak combination of
3038 OpenSSL+DH_params+TLSSNI.
3040 PP/39 Disable SSLv2 by default in OpenSSL support.
3046 PP/01 Solaris build fix for Oracle's LDAP libraries.
3047 Bugzilla 1109, patch from Stephen Usher.
3049 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3051 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3054 TF/02 Fix a couple more cases where we did not log the error message
3055 when unlink() failed. See also change 4.74-TF/03.
3057 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3058 lock up or crash if it happened to be inside a call to libc when it
3059 got a SIGUSR1 from exiwhat.
3061 The SIGUSR1 handler appends the current process status to the process
3062 log which is later printed by exiwhat. It used to use the general
3063 purpose logging code to do this, but several functions it calls are
3064 not safe for signals.
3066 The new output code in the SIGUSR1 handler is specific to the process
3067 log, and simple enough that it's easy to inspect for signal safety.
3068 Removing some special cases also simplifies the general logging code.
3069 Removing the spurious timestamps from the process log simplifies
3072 TF/04 Improved ratelimit ACL condition.
3074 The /noupdate option has been deprecated in favour of /readonly which
3075 has clearer semantics. The /leaky, /strict, and /readonly update modes
3076 are mutually exclusive. The update mode is no longer included in the
3077 database key; it just determines when the database is updated. (This
3078 means that when you upgrade Exim will forget old rate measurements.)
3080 Exim now checks that the per_* options are used with an update mode that
3081 makes sense for the current ACL. For example, when Exim is processing a
3082 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3083 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3084 must specify per_mail/readonly. If you omit the update mode it defaults to
3085 /leaky where that makes sense (as before) or /readonly where required.
3087 The /noupdate option is now undocumented but still supported for
3088 backwards compatibility. It is equivalent to /readonly except that in
3089 ACLs where /readonly is required you may specify /leaky/noupdate or
3090 /strict/noupdate which are treated the same as /readonly.
3092 A useful new feature is the /count= option. This is a generalization
3093 of the per_byte option, so that you can measure the throughput of other
3094 aggregate values. For example, the per_byte option is now equivalent
3095 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3097 The per_rcpt option has been generalized using the /count= mechanism
3098 (though it's more complicated than the per_byte equivalence). When it is
3099 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3100 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3101 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3102 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3103 Note that using per_rcpt with a non-readonly update mode in more than
3104 one ACL will cause the recipients to be double-counted. (The per_mail
3105 and per_byte options don't have this problem.)
3107 The handling of very low rates has changed slightly. If the computed rate
3108 is less than the event's count (usually one) then this event is the first
3109 after a long gap. In this case the rate is set to the same as this event's
3110 count, so that the first message of a spam run is counted properly.
3112 The major new feature is a mechanism for counting the rate of unique
3113 events. The new per_addr option counts the number of different
3114 recipients that someone has sent messages to in the last time period. It
3115 behaves like per_rcpt if all the recipient addresses are different, but
3116 duplicate recipient addresses do not increase the measured rate. Like
3117 the /count= option this is a general mechanism, so the per_addr option
3118 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3119 example, measure the rate that a client uses different sender addresses
3120 with the options per_mail/unique=$sender_address. There are further
3121 details in the main documentation.
3123 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3125 TF/06 Removed a few PCRE remnants.
3127 TF/07 Automatically extract Exim's version number from tags in the git
3128 repository when doing development or release builds.
3130 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3131 Bugzilla 879. Patch from Paul Fisher.
3133 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3134 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3137 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3139 PP/05 Variable $av_failed, true if the AV scanner deferred.
3140 Bugzilla 1078. Patch from John Horne.
3142 PP/06 Stop make process more reliably on build failure.
3143 Bugzilla 1087. Patch from Heiko Schlittermann.
3145 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3146 Bugzilla 1089. Patch from Heiko Schlittermann.
3148 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3149 Bugzilla 1131. Patch from Holger Weiß.
3151 PP/09 Handle IPv6 addresses with SPF.
3152 Bugzilla 860. Patch from Wolfgang Breyha.
3154 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3156 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3159 PP/11 match_* no longer expand right-hand-side by default.
3160 New compile-time build option, EXPAND_LISTMATCH_RHS.
3161 New expansion conditions, "inlist", "inlisti".
3163 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3165 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3167 PP/14 fix log_write() format string regression from TF/03.
3168 Bugzilla 1152. Patch from Dmitry Isaikin.
3174 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3176 PP/02 Harmonised TLS library version reporting; only show if debugging.
3177 Layout now matches that introduced for other libraries in 4.74 PP/03.
3179 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3181 PP/04 New "dns_use_edns0" global option.
3183 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3186 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3187 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3189 TK/01 Updated PolarSSL code to 0.14.2.
3190 Bugzilla 1097. Patch from Andreas Metzler.
3192 PP/07 Catch divide-by-zero in ${eval:...}.
3193 Fixes bugzilla 1102.
3195 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3198 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3199 format-string attack -- SECURITY: remote arbitrary code execution.
3201 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3202 time unintentionally subject to list matching rules, letting the header
3203 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3204 arbitrary string expansion). This allowed for information disclosure.
3206 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3207 INT_MIN/-1 -- value coerced to INT_MAX.
3213 NM/01 Workaround for PCRE version dependency in version reporting
3216 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3217 This fixes portability to compilers other than gcc, notably
3218 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3220 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3221 makefiles for portability to HP-UX and POSIX correctness.
3223 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3224 Also via indented variable definition in the Makefile.
3225 (Debugging by Oliver Heesakkers).
3227 PP/02 Restore caching of spamd results with expanded spamd_address.
3228 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3230 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3231 Improves build reliability. Fix from: Frank Elsner
3233 NM/02 Fix wide character breakage in the rfc2047 coding
3234 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3236 NM/03 Allow underscore in dnslist lookups
3237 Fixes bug 1026. Patch from Graeme Fowler
3239 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3240 Code patches from Adam Ciarcinski of NetBSD.
3242 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3245 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3246 is logged, to avoid truncation. Patch from John Horne.
3248 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3249 Patch from Jakob Hirsch.
3251 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3252 SQL string expansion failure details.
3253 Patch from Andrey Oktyabrski.
3255 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3256 Patch from Simon Arlott.
3258 PP/09 New lookups functionality failed to compile on old gcc which rejects
3259 extern declarations in function scope.
3260 Patch from Oliver Fleischmann
3262 PP/10 Use sig_atomic_t for flags set from signal handlers.
3263 Check getgroups() return and improve debugging.
3264 Fixed developed for diagnosis in bug 927 (which turned out to be
3267 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3268 Patch from Mark Zealey.
3270 PP/12 Bugzilla 1056: Improved spamd server selection.
3271 Patch from Mark Zealey.
3273 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3274 Based on patch from Heiko Schlittermann.
3276 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3277 Patch from Uwe Doering, sign-off by Michael Haardt.
3279 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3280 variable declaration deep within a block. Bug and patch from
3283 PP/15 lookups-Makefile IRIX compatibility coercion.
3285 PP/16 Make DISABLE_DKIM build knob functional.
3287 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3288 Patch by Simon Arlott
3290 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3291 variable argument macros. Our copy now differs from upstream.
3297 TF/01 Failure to get a lock on a hints database can have serious
3298 consequences so log it to the panic log.
3300 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3301 controlled using the smtp_confirmation log selector.
3303 TF/03 Include the error message when we fail to unlink a spool file.
3305 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3306 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3307 for maintaining out-of-tree patches for some time.
3309 PP/01 Bugzilla 139: Documentation and portability issues.
3310 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3311 Handle per-OS dynamic-module compilation flags.
3313 PP/02 Let /dev/null have normal permissions.
3314 The 4.73 fixes were a little too stringent and complained about the
3315 permissions on /dev/null. Exempt it from some checks.
3316 Reported by Andreas M. Kirchwitz.
3318 PP/03 Report version information for many libraries, including
3319 Exim version information for dynamically loaded libraries. Created
3320 version.h, now support a version extension string for distributors
3321 who patch heavily. Dynamic module ABI change.
3323 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3324 privilege escalation vulnerability whereby the Exim run-time user
3325 can cause root to append content of the attacker's choosing to
3328 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3331 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3332 If dropping privileges for untrusted macros, we disabled normal logging
3333 on the basis that it would fail; for the Exim run-time user, this is not
3334 the case, and it resulted in successful deliveries going unlogged.
3335 Fixed. Reported by Andreas Metzler.
3341 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3342 only prepend for the Resent-* case. Fixes regression introduced in
3343 Exim 4.70 by NM/22 for Bugzilla 607.
3345 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3346 increasing numbers of administrators be bitten by this.
3348 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3350 PP/03 Bugzilla 994: added openssl_options main configuration option.
3352 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3354 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3356 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3358 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3359 without a peer certificate, leading to a segfault because of an
3360 assumption that peers always have certificates. Be a little more
3361 paranoid. Problem reported by Martin Tscholak.
3363 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3364 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3365 NB: ClamAV planning to remove STREAM in "middle of 2010".
3366 CL also introduces -bmalware, various -d+acl logging additions and
3367 more caution in buffer sizes.
3369 PP/09 Implemented reverse_ip expansion operator.
3371 PP/10 Bugzilla 937: provide a "debug" ACL control.
3373 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3375 PP/12 Bugzilla 973: Implement --version.
3377 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3379 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3381 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3383 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3384 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3385 ignore trailing whitespace.
3387 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3389 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3392 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3393 Notification from Dr Andrew Aitchison.
3395 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3396 ExtendedDetectionInfo response format.
3397 Notification from John Horne.
3399 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3402 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3403 XSL and documented dependency on system catalogs, with examples of how
3406 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3409 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3410 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3411 configuration file which is writeable by the Exim user or group.
3413 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3414 of configuration files to cover files specified with the -C option if
3415 they are going to be used with root privileges, not just the default
3418 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3419 option (effectively making it always true).
3421 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3422 files to be used while preserving root privileges.
3424 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3425 that rogue child processes cannot use them.
3427 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3428 run-time user, instead of root.
3430 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3431 Exim run-time user without dropping privileges.
3433 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3434 result string, instead of calling string_vformat() twice with the same
3437 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3438 for other users. Others should always drop root privileges if they use
3439 -C on the command line, even for a whitelisted configure file.
3441 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3443 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3449 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3450 $data_path, and $header_path variables; fixed documentation bugs and
3453 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3454 exipick to access non-standard spools, including the "frozen" queue
3457 NM/01 Bugzilla 965: Support mysql stored procedures.
3458 Patch from Alain Williams
3460 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3462 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3463 Patch from Andreas Metzler
3465 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3466 Patch from Kirill Miazine
3468 NM/05 Bugzilla 671: Added umask to procmail example.
3470 JJ/03 installed exipick 20100323.0, fixing doc bug
3472 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3473 directory. Notification and patch from Dan Rosenberg.
3475 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3477 TK/02 Improve log output when DKIM signing operation fails.
3479 MH/01 Treat the transport option dkim_domain as a colon separated
3480 list, not as a single string, and sign the message with each element,
3481 omitting multiple occurences of the same signer.
3483 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3484 Bugzilla 985, 986. Patch by Simon Arlott
3486 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3487 Patch by Simon Arlott
3489 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3490 MBX locking. Notification from Dan Rosenberg.
3496 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3498 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3500 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3502 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3504 NM/04 Bugzilla 915: Flag broken perl installation during build.
3510 TK/01 Added patch by Johannes Berg that expands the main option
3511 "spamd_address" if it starts with a dollar sign.
3513 TK/02 Write list of recipients to X-Envelope-Sender header when building
3514 the mbox-format spool file for content scanning (suggested by Jakob
3517 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3518 (http://www.dcc-servers.net/) support via dccifd. Activated by
3519 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3521 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3522 by Mark Daniel Reidel <mr@df.eu>.
3524 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3525 When building exim an external PCRE library is now needed -
3526 PCRE is a system library on the majority of modern systems.
3527 See entry on PCRE_LIBS in EDITME file.
3529 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3530 conversation. Added nologin parameter to request.
3531 Patch contributed by Kirill Miazine.
3533 TF/01 Do not log submission mode rewrites if they do not change the address.
3535 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3537 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3538 log files in place. Contributed by Roberto Lima.
3540 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3542 TF/03 Bugzilla 615: When checking the local_parts router precondition
3543 after a local_part_suffix or local_part_prefix option, Exim now
3544 does not use the address's named list lookup cache, since this
3545 contains cached lookups for the whole local part.
3547 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3548 Robert Millan. Documentation is in experimental-spec.txt.
3550 TF/04 Bugzilla 668: Fix parallel build (make -j).
3552 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3554 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3555 Patch provided by Jan Srzednicki.
3557 TF/05 Leading white space used to be stripped from $spam_report which
3558 wrecked the formatting. Now it is preserved.
3560 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3561 that they are available at delivery time.
3563 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3565 TF/08 TLS error reporting now respects the incoming_interface and
3566 incoming_port log selectors.
3568 TF/09 Produce a more useful error message if an SMTP transport's hosts
3569 setting expands to an empty string.
3571 NM/06 Bugzilla 744: EXPN did not work under TLS.
3572 Patch provided by Phil Pennock.
3574 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3575 Patch provided by Richard Godbee.
3577 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3578 acl_smtp_notquit, added index entry.
3580 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3581 Patch provided by Eugene Bujak.
3583 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3584 accept(). Patch provided by Maxim Dounin.
3586 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3587 Patch provided by Phil Pennock.
3589 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3591 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3592 Patch provided by Brad "anomie" Jorsch.
3594 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3595 Patch provided by Dean Brooks.
3597 TK/05 Add native DKIM support (does not depend on external libraries).
3599 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3600 Patch provided by Graeme Fowler.
3602 NM/16 Bugzilla 851: Documentation example syntax fix.
3604 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3606 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3609 NM/19 Bugzilla 745: TLS version reporting.
3610 Patch provided by Phil Pennock.
3612 NM/20 Bugzilla 167: bool: condition support.
3613 Patch provided by Phil Pennock.
3615 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3616 clients. Patch provided by Phil Pennock.
3618 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3619 Patch provided by Brad "anomie" Jorsch.
3621 NM/23 Bugzilla 687: Fix misparses in eximstats.
3622 Patch provided by Heiko Schlittermann.
3624 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3625 Patch provided by Heiko Schlittermann.
3627 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3628 plus update to original patch.
3630 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3632 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3633 Patch provided by David Brownlee.
3635 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3637 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3639 NM/30 Bugzilla 888: TLS documentation bugfixes.
3641 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3643 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3644 Unlike the original bugzilla I have changed all shell scripts in src tree.
3646 NM/33 Bugzilla 898: Transport filter timeout fix.
3647 Patch by Todd Rinaldo.
3649 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3650 Patch by Serge Demonchaux.
3652 NM/35 Bugzilla 39: Base64 decode bug fixes.
3653 Patch by Jakob Hirsch.
3655 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3657 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3659 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3661 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3667 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3668 ALT-N's libdkim that I have put here:
3669 http://duncanthrax.net/exim-experimental/
3671 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3672 were called 'true' and it seems that C99 defines that as a reserved
3673 keyword to be used with 'bool' variable types. That means you could
3674 not include C99-style headers which use bools without triggering
3675 build errors in sieve.c.
3677 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3678 as mailq or other aliases. Changed the --help handling significantly
3679 to do whats expected. exim_usage() emits usage/help information.
3681 SC/01 Added the -bylocaldomain option to eximstats.
3683 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3685 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3687 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3693 PH/01 Another patch from the Sieve maintainer.
3695 PH/02 When an IPv6 address is converted to a string for single-key lookup
3696 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3697 dots are used instead of colons so that keys in lsearch files need not
3698 contain colons. This was done some time before quoting was made available
3699 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3700 (notated using the quote facility) so as to distinguish them from IPv4
3701 keys. This meant that lookups for IP addresses in host lists did not work
3702 for iplsearch lookups.
3704 This has been fixed by arranging for IPv6 addresses to be expressed with
3705 colons if the lookup type is iplsearch. This is not incompatible, because
3706 previously such lookups could never work.
3708 The situation is now rather anomalous, since one *can* have colons in
3709 ordinary lsearch keys. However, making the change in all cases is
3710 incompatible and would probably break a number of configurations.
3712 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3715 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3716 conversion specification without a maximum field width, thereby enabling
3717 a rogue spamd server to cause a buffer overflow. While nobody in their
3718 right mind would setup Exim to query an untrusted spamd server, an
3719 attacker that gains access to a server running spamd could potentially
3720 exploit this vulnerability to run arbitrary code as the Exim user.
3722 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3723 $primary_hostname instead of what libspf2 thinks the hosts name is.
3725 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3726 a directory entry by the name of the lookup key. Previously, if a
3727 symlink pointed to a non-existing file or a file in a directory that
3728 Exim lacked permissions to read, a lookup for a key matching that
3729 symlink would fail. Now it is enough that a matching directory entry
3730 exists, symlink or not. (Bugzilla 503.)
3732 PH/03 The body_linecount and body_zerocount variables are now exported in the
3735 PH/04 Added the $dnslist_matched variable.
3737 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3738 This means they are set thereafter only if the connection becomes
3741 PH/06 Added the client_condition to authenticators so that some can be skipped
3742 by clients under certain conditions.
3744 PH/07 The error message for a badly-placed control=no_multiline_responses left
3745 "_responses" off the end of the name.
3747 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3749 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3750 (without spaces) instead of just copying the configuration text.
3752 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3754 PH/11 Added $max_received_linelength.
3756 PH/12 Added +ignore_defer and +include_defer to host lists.
3758 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3759 way in which PCRE > 7.0 is built.
3761 PH/14 Implemented queue_only_load_latch.
3763 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3764 MAIL command. The effect was to mangle the value on 64-bit systems.
3766 PH/16 Another patch from the Sieve maintainer.
3768 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3770 PH/18 If a system quota error occurred while trying to create the file for
3771 a maildir delivery, the message "Mailbox is full" was not appended to the
3772 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3773 only to a quota excession during the actual writing of the file.
3775 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3776 characters?) which causes problems in log lines. The DN values are now
3777 passed through string_printing() before being added to log lines.
3779 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3780 and InterBase are left for another time.)
3782 PH/21 Added message_body_newlines option.
3784 PH/22 Guard against possible overflow in moan_check_errorcopy().
3786 PH/23 POSIX allows open() to be a macro; guard against that.
3788 PH/24 If the recipient of an error message contained an @ in the local part
3789 (suitably quoted, of course), incorrect values were put in $domain and
3790 $local_part during the evaluation of errors_copy.
3796 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3797 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3800 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3801 issue a MAIL command.
3803 PH/02 In an ACL statement such as
3805 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3807 if a client was not listed at all, or was listed with a value other than
3808 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3809 the condition was not true (as it should be), so access was not denied.
3810 The bug was that the ! inversion was incorrectly passed on to the second
3811 item. This has been fixed.
3813 PH/03 Added additional dnslists conditions == and =& which are different from
3814 = and & when the dns lookup returns more than one IP address.
3816 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3817 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3819 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3820 FSYNC, which compiles an option called disable_fsync that allows for
3821 bypassing fsync(). The documentation is heavily laced with warnings.
3823 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3825 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3826 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3827 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3828 including adding "make clean"; (3) Added -fPIC when compiling the test
3829 dynamically loaded module, to get rid of a warning.
3831 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3832 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3833 = 0s. The bug is otherwise harmless.
3835 PH/07 There was a bug in the dovecot authenticator such that the value of
3836 $auth1 could be overwritten, and so not correctly preserved, after a
3837 successful authentication. This usually meant that the value preserved by
3838 the server_setid option was incorrect.
3840 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3842 PH/09 Installed PCRE release 7.0.
3844 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3845 run for batched SMTP input. It is now run at the start of every message
3846 in the batch. While fixing this I discovered that the process information
3847 (output by running exiwhat) was not always getting set for -bs and -bS
3848 input. This is fixed, and it now also says "batched" for BSMTP.
3850 PH/11 Added control=no_pipelining.
3852 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3853 patch, slightly modified), and move the expansion of helo_data till after
3854 the connection is made in the smtp transport (so it can use these
3857 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3859 PH/14 Added log_selector = +pid.
3861 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3863 PH/16 Add ${if forany and ${if forall.
3865 PH/17 Added dsn_from option to vary the From: line in DSNs.
3867 PH/18 Flush SMTP output before performing a callout, unless control =
3868 no_callout_flush is set.
3870 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3871 was true (the default) a successful delivery failed to delete the retry
3872 item, thus causing premature timeout of the address. The bug is now
3875 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3877 PH/21 Long custom messages for fakedefer and fakereject are now split up
3878 into multiline responses in the same way that messages for "deny" and
3879 other ACL rejections are.
3881 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3882 with slight modification.
3884 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3885 draft, changing the syntax and factoring some duplicate code".
3887 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3888 for deliveries of the second and subsequent messages over the same SMTP
3891 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3892 ${reduce, with only minor "tidies".
3894 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3896 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3897 expansion side effects.
3899 PH/27 When a message times out after an over-quota error from an Exim-imposed
3900 quota, the bounce message says "mailbox is full". This message was not
3901 being given when it was a system quota that was exceeded. It now should
3904 MH/03 Made $recipients available in local_scan(). local_scan() already has
3905 better access to the recipient list through recipients_list[], but
3906 $recipients can be useful in postmaster-provided expansion strings.
3908 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3909 in the case of a MAIL command with additional options following the
3910 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3911 were accidentally chopped off.
3913 PH/29 SMTP synchronization checks are implemented when a command is read -
3914 there is a check that no more input is waiting when there shouldn't be
3915 any. However, for some commands, a delay in an ACL can mean that it is
3916 some time before the response is written. In this time, more input might
3917 arrive, invalidly. So now there are extra checks after an ACL has run for
3918 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3919 pipelining has not been advertised.
3921 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3923 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3924 returned in the SMTP response when smtp_return_error_details was false.
3925 This has been fixed.
3927 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3928 instead of the C I/O that was originally supplied, because problems were
3929 reported on Solaris.
3931 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3932 Exim which did not show up earlier: it was assuming that a call to
3933 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3934 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3935 was a macro that became an assignment, so it seemed to work. This has
3936 changed to a proper function call with a void return, hence the compile
3937 error. Exim's code has been fixed.
3939 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3942 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3944 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3946 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3947 because it turns out to be too big or there is a timeout) there is no
3948 "Completed" line in the log. When some messages of this type were
3949 selected by exigrep, they were listed as "not completed". Others were
3950 picked up by some special patterns. I have improved the selection
3951 criteria to be more general.
3953 PH/38 The host_find_failed option in the manualroute router can now be set
3954 to "ignore", to completely ignore a host whose IP address cannot be
3955 found. If all hosts are ignored, the behaviour is controlled by the new
3956 host_all_ignored option.
3958 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3959 homing or because of multiple MX records with /mx) generated more than
3960 one IP address, and the following item turned out to be the local host,
3961 all the secondary addresses of the first item were incorrectly removed
3962 from the list, along with the local host and any following hosts (which
3963 is what is supposed to happen).
3965 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3966 whoever called Exim into the -H file. In the case of the daemon it was
3967 behaving confusingly. When first started, it used values for whoever
3968 started the daemon, but after a SIGHUP it used the Exim user (because it
3969 calls itself on a restart). I have changed the code so that it now always
3972 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3973 message are rejected with the same error (e.g. no authentication or bad
3974 sender address), and a DATA command is nevertheless sent (as can happen
3975 with PIPELINING or a stupid MUA), the error message that was given to the
3976 RCPT commands is included in the rejection of the DATA command. This is
3977 intended to be helpful for MUAs that show only the final error to their
3980 PH/42 Another patch from the Sieve maintainer.
3982 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3983 Eximstats - Fixed some broken HTML links and added missing column headers
3985 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3986 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3988 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3990 PH/43 Yet another patch from the Sieve maintainer.
3992 PH/44 I found a way to check for a TCP/IP connection going away before sending
3993 the response to the final '.' that terminates a message, but only in the
3994 case where the client has not sent further data following the '.'
3995 (unfortunately, this is allowed). However, in many cases there won't be
3996 any further data because there won't be any more messages to send. A call
3997 to select() can be used: if it shows that the input is "ready", there is
3998 either input waiting, or the socket has been closed. An attempt to read
3999 the next input character can distinguish the two cases. Previously, Exim
4000 would have sent an OK response which the client would never have see.
4001 This could lead to message repetition. This fix should cure that, at
4002 least in a lot of common cases.
4004 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4005 advertised in response to EHLO.
4011 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4012 fixed by 4.65/MH/01 (is this a record?) are fixed:
4014 (i) An empty string was always treated as zero by the numeric comparison
4015 operators. This behaviour has been restored.
4017 (ii) It is documented that the numeric comparison operators always treat
4018 their arguments as decimal numbers. This was broken in that numbers
4019 starting with 0 were being interpreted as octal.
4021 While fixing these problems I realized that there was another issue that
4022 hadn't been noticed. Values of message_size_limit (both the global option
4023 and the transport option) were treated as octal if they started with 0.
4024 The documentation was vague. These values are now always treated as
4025 decimal, and I will make that clear in the documentation.
4031 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4032 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4035 MH/01 Don't check that the operands of numeric comparison operators are
4036 integers when their expansion is in "skipping" mode (fixes bug
4037 introduced by 4.64-PH/07).
4039 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4040 child addresses, Exim now panics and dies. Previously, because the count
4041 is held in a short int, deliveries were likely to be lost. As such a
4042 large number of recipients for a single message is ridiculous
4043 (performance will be very, very poor), I have chosen to impose a limit
4044 rather than extend the field.
4050 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4051 leftover -K file (the existence of which was triggered by #402).
4052 While we were at it, introduced process PID as part of the -K
4053 filename. This should rule out race conditions when creating
4056 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4057 processing considerably. Previous code took too long for large mails,
4058 triggering a timeout which in turn triggers #401.
4060 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4061 in the DK code in transports.c. sendfile() is not really portable,
4062 hence the _LINUX specificness.
4064 TF/01 In the add_headers option to the mail command in an Exim filter,
4065 there was a bug that Exim would claim a syntax error in any
4066 header after the first one which had an odd number of characters
4069 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4070 callout verification, Exim cached a "reject" for the entire domain. This
4071 is correct for most verifications, but it is not correct for a recipient
4072 verification with use_sender or use_postmaster set, because in that case
4073 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4074 case of MAIL FROM:<> rejection from other early rejections (e.g.
4075 rejection of HELO). When verifying a recipient using a non-null MAIL
4076 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4077 Whatever the result of the callout, the value of the domain cache is
4078 left unchanged (for any other kind of callout, getting as far as trying
4079 RCPT means that the domain itself is ok).
4081 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4084 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4085 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4086 session, the daemon ignores ECONNECT errors and logs others; it now
4087 ignores EPIPE as well.
4089 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4090 (quoted-printable decoding).
4092 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4093 later the small subsequent patch to fix an introduced bug.
4095 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4097 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4099 PH/08 An error is now given if message_size_limit is specified negative.
4101 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4102 to be given (somewhat) arbitrary names.
4104 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4107 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4108 miscellaneous code fixes
4110 PH/10 Added the log_reject_target ACL modifier to specify where to log
4113 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4114 hostname. This is wrong, because it relates to the incoming message (and
4115 probably the interface on which it is arriving) and not to the outgoing
4116 callout (which could be using a different interface). This has been
4117 changed to use the value of the helo_data option from the smtp transport
4118 instead - this is what is used when a message is actually being sent. If
4119 there is no remote transport (possible with a router that sets up host
4120 addresses), $smtp_active_hostname is used.
4122 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4123 tweaks were necessary in order to get it to work (see also 21 below):
4124 (a) The code assumed that strncpy() returns a negative number on buffer
4125 overflow, which isn't the case. Replaced with Exim's string_format()
4127 (b) There were several signed/unsigned issues. I just did the minimum
4128 hacking in of casts. There is scope for a larger refactoring.
4129 (c) The code used strcasecmp() which is not a standard C function.
4130 Replaced with Exim's strcmpic() function.
4131 (d) The code set only $1; it now sets $auth1 as well.
4132 (e) A simple test gave the error "authentication client didn't specify
4133 service in request". It would seem that Dovecot has changed its
4134 interface. Fortunately there's a specification; I followed it and
4135 changed what the client sends and it appears to be working now.
4137 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4140 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4141 address A is aliased to B and C, where B exists and C does not. Without
4142 -v the output is "A verified" because verification stops after a
4143 successful redirection if more than one address is generated. However,
4144 with -v the child addresses are also verified. Exim was outputting "A
4145 failed to verify" and then showing the successful verification for C,
4146 with its parentage. It now outputs "B failed to verify", showing B's
4147 parentage before showing the successful verification of C.
4149 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4150 look up a TXT record in a specific list after matching in a combined
4153 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4154 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4155 they consult the DNS. I had assumed they would set it the way they
4156 wanted; and indeed my experiments on Linux seem to show that in some
4157 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4158 To be on the safe side, however, I have now made the interface to
4159 host_find_byname() similar to host_find_bydns(), with an argument
4160 containing the DNS resolver options. The host_find_byname() function now
4161 sets these options at its start, just as host_find_bydns() does. The smtp
4162 transport options dns_qualify_single and dns_search_parents are passed to
4163 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4164 of host_find_byname() use the default settings of RES_DEFNAMES
4165 (qualify_single) but not RES_DNSRCH (search_parents).
4167 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4168 spool_read_header() do less string testing, by means of a preliminary
4169 switch on the second character of optional "-foo" lines. (This is
4170 overdue, caused by the large number of possibilities that now exist.
4171 Originally there were few.) While I was there, I also converted the
4172 str(n)cmp tests so they don't re-test the leading "-" and the first
4173 character, in the hope this might squeeze out yet more improvement.
4175 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4176 flag allowing group syntax was set by the header_syntax check but not
4177 turned off, possible causing trouble later; (2) The flag was not being
4178 set at all for the header_verify test, causing "group"-style headers to
4179 be rejected. I have now set it in this case, and also caused header_
4180 verify to ignore an empty address taken from a group. While doing this, I
4181 came across some other cases where the code for allowing group syntax
4182 while scanning a header line wasn't quite right (mostly, not resetting
4183 the flag correctly in the right place). These bugs could have caused
4184 trouble for malformed header lines. I hope it is now all correct.
4186 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4187 with the "reply" argument non-NULL. The code, however (which originally
4188 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4189 but it didn't always do it. This confused somebody who was copying the
4190 code for some other use. I have removed all the tests.
4192 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4193 feature that was used to support insecure browsers during the U.S. crypto
4194 embargo. It requires special client support, and Exim is probably the
4195 only MTA that supported it -- and would never use it because real RSA is
4196 always available. This code has been removed, because it had the bad
4197 effect of slowing Exim down by computing (never used) parameters for the
4198 RSA_EXPORT functionality.
4200 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4201 authenticator to fail if there's a tab character in the incoming data
4202 (there should never be unless someone is messing about, as it's supposed
4203 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4204 if the connection is using TLS or if the remote IP is the same as the
4205 local IP, and the "valid-client-cert option" if a client certificate has
4208 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4209 authenticators. This can be used for authorization after authentication
4210 succeeds. (In the case of plaintext, it servers for both authentication
4213 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4214 if any retry times were supplied.
4216 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4217 connection, where there is no client IP address to check. In this
4218 situation, the verify now always succeeds.
4220 PH/25 Applied John Jetmore's -Mset patch.
4222 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4224 PH/27 In a string expansion for a processed (not raw) header when multiple
4225 headers of the same name were present, leading whitespace was being
4226 removed from all of them, but trailing whitespace was being removed only
4227 from the last one. Now trailing whitespace is removed from each header
4228 before concatenation. Completely empty headers in a concatenation (as
4229 before) are ignored.
4231 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4232 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4234 PH/29 [Removed. This was a change that I later backed out, and forgot to
4235 correct the ChangeLog entry (that I had efficiently created) before
4236 committing the later change.]
4238 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4239 address errors (4xx response to RCPT) over the same connection as other
4240 messages routed to the same hosts. Such deliveries are always "forced",
4241 so retry times are not inspected. This resulted in far too many retries
4242 for the affected addresses. The effect occurred only when there were more
4243 hosts than the hosts_max_try setting in the smtp transport when it had
4244 the 4xx errors. Those hosts that it had tried were not added to the list
4245 of hosts for which the message was waiting, so if all were tried, there
4246 was no problem. Two fixes have been applied:
4248 (i) If there are any address or message errors in an SMTP delivery, none
4249 of the hosts (tried or untried) are now added to the list of hosts
4250 for which the message is waiting, so the message should not be a
4251 candidate for sending over the same connection that was used for a
4252 successful delivery of some other message. This seems entirely
4253 reasonable: after all the message is NOT "waiting for some host".
4254 This is so "obvious" that I'm not sure why it wasn't done
4255 previously. Hope I haven't missed anything, but it can't do any
4256 harm, as the worst effect is to miss an optimization.
4258 (ii) If, despite (i), such a delivery is accidentally attempted, the
4259 routing retry time is respected, so at least it doesn't keep
4260 hammering the server.
4262 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4263 in ${readsocket because some servers need this prod.
4265 PH/32 Added some extra debug output when updating a wait-xxx database.
4267 PH/33 The hint "could be header name not terminated by colon", which has been
4268 given for certain expansion errors for a long time, was not being given
4269 for the ${if def:h_colon_omitted{... case.
4271 PH/34 The spec says: "With one important exception, whenever a domain list is
4272 being scanned, $domain contains the subject domain." There was at least
4273 one case where this was not true.
4275 PH/35 The error "getsockname() failed: connection reset by peer" was being
4276 written to the panic log as well as the main log, but it isn't really
4277 panic-worthy as it just means the connection died rather early on. I have
4278 removed the panic log writing for the ECONNRESET error when getsockname()
4281 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4282 runs only) independently of the message's sender address. This meant
4283 that, if the 4xx error was in fact related to the sender, a different
4284 message to the same recipient with a different sender could confuse
4285 things. In particular, this can happen when sending to a greylisting
4286 server, but other circumstances could also provoke similar problems.
4287 I have changed the default so that the retry time for these errors is now
4288 based a combination of the sender and recipient addresses. This change
4289 can be overridden by setting address_retry_include_sender=false in the
4292 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4293 remote server are returned as part of bounce messages. This was not
4294 happening for LMTP over a pipe (the lmtp transport), but now it is the
4295 same for both kinds of LMTP.
4297 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4298 in header lines that were in fact CNAMEs. This is no longer the case.
4300 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4301 and queue runs started by the daemon processed all messages. This has
4302 been fixed so that -R and -S can now usefully be given with -q<time>.
4304 PH/40 Import PCRE release 6.7 (fixes some bugs).
4306 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4308 PH/42 Give an error if -q is specified more than once.
4310 PH/43 Renamed the variables $interface_address and $interface_port as
4311 $received_ip_address and $received_port, to make it clear that these
4312 values apply to message reception, and not to the outgoing interface when
4313 a message is delivered. (The old names remain recognized, of course.)
4315 PH/44 There was no timeout on the connect() call when using a Unix domain
4316 socket in the ${readsocket expansion. There now is.
4318 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4319 be meaningful with "accept".
4321 SC/01 Eximstats V1.43
4322 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4324 SC/02 Eximstats V1.44
4325 Use a glob alias rather than an array ref in the generated
4326 parser. This improves both readability and performance.
4328 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4329 Collect SpamAssassin and rejection statistics.
4330 Don't display local sender or destination tables unless
4331 there is data to show.
4332 Added average volumes into the top table text output.
4334 SC/04 Eximstats V1.46
4335 Collect data on the number of addresses (recipients)
4336 as well as the number of messages.
4338 SC/05 Eximstats V1.47
4339 Added 'Message too big' to the list of mail rejection
4340 reasons (thanks to Marco Gaiarin).
4342 SC/06 Eximstats V1.48
4343 Mainlog lines which have GMT offsets and are too short to
4344 have a flag are now skipped.
4346 SC/07 Eximstats V1.49 (Alain Williams)
4347 Added the -emptyok flag.
4349 SC/08 Eximstats V1.50
4350 Fixes for obtaining the IP address from reject messages.
4352 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4353 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4354 whitespace changes from 4.64-PH/27
4356 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4359 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4360 are found, allow negative numbers in numeric criteria)
4362 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4364 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4365 to match changes made in 4.64-PH/43
4367 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4369 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4370 "domain" is actually an IP address, and force a failure. This locks out
4371 those revolvers/nameservers that support "A-for-A" lookups, in
4372 contravention of the specifications.
4374 PH/48 When a host name was looked up from an IP address, and the subsequent
4375 forward lookup of the name timed out, the host name was left in
4376 $sender_host_name, contrary to the specification.
4378 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4379 restricted to single-key lookups, Exim was not diagnosing an error if
4380 * or *@ was used with a query-style lookup.
4382 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4384 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4385 long ago, but noone interested enough thought of it. Let's just say that
4386 the "1.1" means that there are some new functions that weren't there at
4387 some point in the past.
4389 PH/51 Error processing for expansion failure of helo_data from an smtp
4390 transport during callout processing was broken.
4392 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4393 tested/used via the -bh/-bhc/-bs options.
4395 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4396 bug, fixed in subsequent PCRE releases).
4398 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4399 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4401 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4407 SC/01 Use a glob alias rather than an array ref in eximstats generated
4408 parser. This improves both readability and performance.
4410 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4411 Don't display local sender or destination tables in eximstats unless
4412 there is data to show.
4413 Added average volumes into the eximstats top table text output.
4415 SC/03 Collect data on the number of addresses (recipients) as well
4416 as the number of messages in eximstats.
4418 TF/01 Correct an error in the documentation for the redirect router. Exim
4419 does (usually) call initgroups() when daemonizing.
4421 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4422 with consistent privilege compared to when running as a daemon.
4424 TF/03 Note in the spec that $authenticated_id is not set for local
4425 submissions from trusted users.
4427 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4428 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4430 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4431 by adding some example configuration directives to the default
4432 configuration file. A little bit of work is required to uncomment the
4433 directives and define how usernames and passwords are checked, but
4434 there is now a framework to start from.
4436 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4437 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4438 without this. I don't know how relevant this is to other LDAP libraries.
4440 PH/02 Add the verb name to the "unknown ACL verb" error.
4442 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4444 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4446 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4447 directory not expanded when it should be if an expanded home directory
4448 was set for the address (which is overridden by the transport).
4450 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4453 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4454 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4455 because it is too late at that time, and has no effect.
4457 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4458 security issue with \' (bugzilla #107). I could not use the
4459 PQescapeStringConn() function, because it needs a PGconn value as one of
4462 PH/08 When testing addresses using -bt, indicate those final addresses that
4463 are duplicates that would not cause an additional delivery. At least one
4464 person was confused, thinking that -bt output corresponded to deliveries.
4465 (Suppressing duplicates isn't a good idea as you lose the information
4466 about possibly different redirections that led to the duplicates.)
4468 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4469 systems where poll() doesn't work, in particular OS X.
4471 PH/10 Added more information to debugging output for retry time not reached.
4473 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4474 operations in malware.c.
4476 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4479 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4480 syslog_duplication set false, log lines that would normally be written
4481 both the the main log and to the reject log were not written to syslog at
4484 PH/14 In the default configuration, change the use of "message" in ACL warn
4485 statements to "add_header".
4487 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4488 not followed by a command (e.g. "seen endif").
4490 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4491 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4494 PH/17 Added extra conditions to the default value of delay_warning_condition
4498 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4499 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4500 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4503 The Auto-Submitted: and various List- headers are standardised, whereas I
4504 don't think Precedence: ever was.
4506 PH/18 Refactored debugging code in route_finduser() to show more information,
4507 in particular, the error code if getpwnam() issues one.
4509 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4510 This is apparently needed in addition to the PH/07 change above to avoid
4511 any possible encoding problems.
4513 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4514 but not after initializing Perl.
4516 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4517 output them only if debugging. By default they are written stderr,
4518 apparently, which is not desirable.
4520 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4523 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4526 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4528 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4529 authenticated or an ident call has been made. Suppress the default
4530 values for $authenticated_id and $authenticated_sender (but permit -oMai
4531 and -oMas) when testing with -bh.
4533 PH/24 Re-jigged the order of the tests in the default configuration so that the
4534 tests for valid domains and recipients precede the DNS black list and CSA
4535 tests, on the grounds that those ones are more expensive.
4537 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4538 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4539 This bug exists in every version of Exim that I still have, right back to
4542 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4543 However, an attempt to turn on case-sensitivity in a regex key by
4544 including (?-i) didn't work because the subject string was already
4545 lowercased, and the effects were non-intuitive. It turns out that a
4546 one-line patch can be used to allow (?-i) to work as expected.
4552 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4553 other effects) broke the use of negated acl sub-conditions.
4555 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4558 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4559 "Deny" causes Exim to reject the incoming connection with a 554 error.
4560 Unfortunately, if there is a major crisis, such as a disk failure,
4561 tcp-wrappers gives "deny", whereas what one would like would be some
4562 kind of temporary error. A kludge has been added to help with this.
4563 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4564 554 error is used if errno is still zero or contains ENOENT (which occurs
4565 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4568 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4570 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4571 errors. Otherwise a message that provokes a temporary error (when other
4572 messages do not) can cause a whole host to time out.
4574 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4575 addresses were routed directly to files or pipes from a redirect router.
4576 File deliveries just didn't batch; pipe deliveries might have suffered
4579 PH/06 A failure to get a lock for a hints database would erroneously always say
4580 "Failed to get write lock", even when it was really a read lock.
4582 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4583 of 0600. This has been changed to use the value of the lockfile_mode
4584 option (which defaults to 0600).
4586 PH/08 Applied small patch from the Sieve maintainer.
4588 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4589 folder from quota calculations, a direct delivery into this folder messed
4590 up the contents of the maildirsize file. This was because the regex was
4591 used only to exclude .Trash (or whatever) when the size of the mailbox
4592 was calculated. There was no check that a delivery was happening into an
4593 excluded directory. This bug has been fixed by ignoring all quota
4594 processing for deliveries into excluded directories.
4596 PH/10 Added the maildirfolder_create_regex option to appendfile.
4602 PH/01 The code for finding all the local interface addresses on a FreeBSD
4603 system running IPv6 was broken. This may well have applied to all BSD
4604 systems, as well as to others that have similar system calls. The broken
4605 code found IPv4 interfaces correctly, but gave incorrect values for the
4606 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4607 that it would not match correctly against @[] and not recognize the IPv6
4610 PH/02 The ipliteral router was not recognizing addresses of the form user@
4611 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4613 PH/03 Added disable_ipv6.
4615 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4616 decoded form, because it is most often used to construct To: headers
4617 lines in autoreplies, and the decoded form may well be syntactically
4618 invalid. However, $reply_address has leading white space removed, and all
4619 newlines turned into spaces so that the autoreply transport does not
4622 PH/05 If group was specified without a user on a router, and no group or user
4623 was specified on a transport, the group from the router was ignored.
4625 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4626 for visible compile-time settings that can be used to change these
4627 numbers, for those that want even more. Backwards compatibility with old
4628 spool files has been maintained. However, going back to a previous Exim
4629 release will lost any variables that are in spool files.
4631 PH/07 Two small changes when running in the test harness: increase delay when
4632 passing a TCP/IP connection to a new process, in case the original
4633 process has to generate a bounce, and remove special handling of
4634 127.0.0.2 (sic), which is no longer necessary.
4636 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4637 be the same on different OS.
4639 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4642 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4643 whether --show-vars was specified or not
4645 JJ/02 exipick: Added support for new ACL variable spool format introduced
4648 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4649 syntactically invalid From: or Reply-to: line, and a filter used this to
4650 generate an autoreply, and therefore failed to obtain an address for the
4651 autoreply, Exim could try to deliver to a non-existent relative file
4652 name, causing unrelated and misleading errors. What now happens is that
4653 it logs this as a hard delivery error, but does not attempt to create a
4656 PH/11 The exinext utility has a -C option for testing purposes, but although
4657 the given file was scanned by exinext itself; it wasn't being passed on
4658 when Exim was called.
4660 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4661 an end-of-file indication when reading a command response.
4663 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4664 compiled. In many other places in Exim, IPv6 addresses are always
4665 recognized, so I have changed this. It also means that IPv4 domain
4666 literals of the form [IPV4:n.n.n.n] are now always recognized.
4668 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4669 used if the router is not running as root, for example, when verifying at
4670 ACL time, or when using -bh. The debugging output from this situation was
4671 non-existent - all you got was a failure to exec. I have made two
4674 (a) Failures to set uid/gid, the current directory, or a process leader
4675 in a subprocess such as that created by queryprogram now generate
4676 suitable debugging output when -d is set.
4678 (b) The queryprogram router detects when it is not running as root,
4679 outputs suitable debugging information if -d is set, and then runs
4680 the subprocess without attempting to change uid/gid.
4682 PH/15 Minor change to Makefile for building test_host (undocumented testing
4685 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4686 additional section of a DNS packet that returns MX or SRV records.
4687 Instead, it always explicitly searches for A/AAAA records. This avoids
4688 major problems that occur when a DNS server includes only records of one
4689 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4690 fixed another bug: if SRV records were looked up and the corresponding
4691 address records were *not* found in the additional section, the port
4692 values from the SRV records were lost.
4694 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4695 using the correct key (the original address) when searching the retry
4696 rules in order to find which one to use for generating the retry hint.
4698 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4699 adding the default one. Similarly, if it contains a Reply-To: header, the
4700 errors_reply_to option, if set, is not used.
4702 PH/19 When calculating a retry time, Exim used to measure the "time since
4703 failure" by looking at the "first failed" field in the retry record. Now
4704 it does not use this if it is later than than the arrival time of the
4705 message. Instead it uses the arrival time. This makes for better
4706 behaviour in cases where some deliveries succeed, thus re-setting the
4707 "first failed" field. An example is a quota failure for a huge message
4708 when small messages continue to be delivered. Without this change, the
4709 "time since failure" will always be short, possible causing more frequent
4710 delivery attempts for the huge message than are intended.
4711 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4713 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4714 $1, $2, $3) because the numerical variables can be reset during some
4715 expansion items (e.g. "match"), thereby losing the authentication data.
4717 PH/21 Make -bV show the size of off_t variables so that the test suite can
4718 decide whether to run tests for quotas > 2G.
4720 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4721 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4722 filecount value is greater than 2G or if a quota value is greater than 2G
4723 on a system where the size of off_t is not greater than 4, a panic error
4726 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4727 never match. The debug and -bh output now contains an explicit error
4728 message indicating a malformed IPv4 address or mask.
4730 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4731 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4732 PH/23 above applies.
4734 PH/25 Do not write to syslog when running in the test harness. The only
4735 occasion when this arises is a failure to open the main or panic logs
4736 (for which there is an explicit test).
4738 PH/26 Added the /no_tell option to "control=freeze".
4740 PH/27 If a host name lookup failed very early in a connection, for example, if
4741 the IP address matched host_lookup and the reverse lookup yielded a name
4742 that did not have a forward lookup, an error message of the form "no IP
4743 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4744 could be logged. Now it outputs the IP address instead of "NULL".
4746 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4747 allows the sender and the authenticated sender to be set when
4748 submitting a message from within Exim. Since child_open_exim() is
4749 documented for local_scan(), the new function should be too.
4751 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4752 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4753 results in an empty string is now treated as unset.
4755 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4757 PH/31 Added sender_verify_fail logging option.
4759 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4760 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4761 tidied the source and removed it altogether.
4763 PH/33 When a queue run was abandoned because the load average was too high, a
4764 log line was always written; now it is written only if the queue_run log
4765 selector is set. In addition, the log line for abandonment now contains
4766 information about the queue run such as the pid. This is always present
4767 in "start" and "stop" lines but was omitted from the "abandon" line.
4769 PH/34 Omit spaces between a header name and the colon in the error message that
4770 is given when verify = headers_syntax fails (if there are lots of them,
4771 the message gets confusing).
4773 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4774 names, as there are now some PTR records that contain slashes. This check
4775 is only to protect against broken name servers that fall over on strange
4776 characters, so the fact that it applies to all lookups doesn't matter.
4778 PH/36 Now that the new test suite is complete, we can remove some of the
4779 special code in Exim that was needed for the old test suite. For example,
4780 sorting DNS records because real resolvers return them in an arbitrary
4781 order. The new test suite's fake resolver always returns records in the
4784 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4785 bounces) except when queue_only is set, to avoid logging races between
4786 the different processes.
4788 PH/38 Panic-die if .include specifies a non-absolute path.
4790 PH/39 A tweak to the "H" retry rule from its user.
4792 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4793 a label. They prevented compilation on older perls.
4795 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4796 a warning to be raised on newish perls.
4798 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4799 on queue. Changes to match documented behaviour of showing count of
4800 messages matching specified criteria.
4802 PH/40 Changed the default ident timeout from 30s to 5s.
4804 PH/41 Added support for the use of login_cap features, on those BSD systems
4805 that have them, for controlling the resources used by pipe deliveries.
4807 PH/42 The content-scanning code uses fopen() to create files in which to put
4808 message data. Previously it was not paying any attention to the mode of
4809 the files. Exim runs with umask(0) because the rest of the code creates
4810 files with open(), and sets the required mode explicitly. Thus, these
4811 files were ending up world-writeable. This was not a big issue, because,
4812 being within the spool directory, they were not world-accessible. I have
4813 created a function called modefopen, which takes an additional mode
4814 argument. It sets umask(777), creates the file, chmods it to the required
4815 mode, then resets the umask. All the relevant calls to fopen() in the
4816 content scanning code have been changed to use this function.
4818 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4819 to 24 hours. This avoids potential overflow problems when processing G
4820 and H retry rules. I suspect nobody ever tinkers with this value.
4822 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4824 PH/45 When the plaintext authenticator is running as a client, the server's
4825 challenges are checked to ensure they are valid base64 strings. By
4826 default, the authentication attempt is cancelled if an invalid string is
4827 received. Setting client_ignore_invalid_base64 true ignores these errors.
4828 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4829 they are received. Thus, the responses can be made to depend on the
4830 challenges. If an invalid string is ignored, an empty string is placed in
4833 PH/46 Messages that are created by the autoreply transport now contains a
4834 References: header, in accordance with RFCs 2822 and 3834.
4836 PH/47 Added authenticated_sender_force to the smtp transport.
4838 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4840 PH/49 Installed latest patch from the Sieve maintainer.
4842 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4843 was also set, the appendfile transport was unnecessarily scanning a
4844 directory of message files (e.g. for maildir delivery) to find the count
4845 of files (along with the size), even though it did not need this
4846 information. It now does the scan only if it needs to find either the
4847 size of the count of files.
4849 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4851 PH/52 Two bugs concerned with error handling when the smtp transport is
4854 (i) Exim was not creating retry information for temporary errors given
4855 for individual recipients after the DATA command when the smtp transport
4856 was used in LMTP mode. This meant that they could be retried too
4857 frequently, and not timed out correctly.
4859 (ii) Exim was setting the flag that allows error details to be returned
4860 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4861 recipients that were returned after the DATA command.
4863 PH/53 This is related to PH/52, but is more general: for any failing address,
4864 when detailed error information was permitted to be returned to the
4865 sender, but the error was temporary, then after the final timeout, only
4866 "retry timeout exceeded" was returned. Now it returns the full error as
4867 well as "retry timeout exceeded".
4869 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4870 do this, and (what is worse) MTAs that accept it.
4872 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4873 will now be deprecated.
4875 PH/56 New os.c-cygwin from the Cygwin maintainer.
4877 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4878 formats (previously only available in exim formats via -bpr, -bpru,
4879 and -bpra. Now also available in native and exiqgrep formats)
4881 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4882 with very large, slow to parse queues
4884 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4886 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4888 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4889 responses to authentication challenges, though it was showing the
4890 challenges; (ii) I've removed the CR characters from the debug output for
4893 PH/58 Allow for the insertion of a newline as well as a space when a string
4894 is turned into more than one encoded-word during RFC 2047 encoding. The
4895 Sieve code now uses this.
4897 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4898 data_4xx, lost_connection, tls_required.
4900 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4901 message was being sent as an SMTP response.
4903 PH/61 Add -l and -k options to exicyclog.
4905 PH/62 When verifying, if an address was redirected to one new address, so that
4906 verification continued, and the new address failed or deferred after
4907 having set something in $address_data, the value of $address_data was not
4908 passed back to the ACL. This was different to the case when no
4909 redirection occurred. The value is now passed back in both cases.
4911 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4912 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4913 use login_cap.h, so on its own it isn't the distinguishing feature. The
4914 new name refers directly to the setclassresources() function.
4916 PH/65 Added configuration files for NetBSD3.
4918 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4920 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4921 is preferred over IPv4.
4923 PH/68 The bounce_return_message and bounce_return_body options were not being
4924 honoured for bounces generated during the reception of non-SMTP messages.
4925 In particular, this applied to messages rejected by the ACL. This bug has
4926 been fixed. However, if bounce_return_message is true and bounce_return_
4927 body is false, the headers that are returned for a non-SMTP message
4928 include only those that have been read before the error was detected.
4929 (In the case of an ACL rejection, they have all been read.)
4931 PH/69 The HTML version of the specification is now built in a directory called
4932 spec_html instead of spec.html, because the latter looks like a path with
4933 a MIME-type, and this confuses some software.
4935 PH/70 Catch two compiler warnings in sieve.c.
4937 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4938 function verify_get_ident() calls ip_connect() to connect a socket, but
4939 if the "connect()" function timed out, ip_connect() used to close the
4940 socket. However, verify_get_ident() also closes the socket later, and in
4941 between Exim writes to the log, which may get opened at this point. When
4942 the socket was closed in ip_connect(), the log could get the same file
4943 descriptor number as the socket. This naturally causes chaos. The fix is
4944 not to close the socket in ip_connect(); the socket should be closed by
4945 the function that creates it. There was only one place in the code where
4946 this was missing, in the iplookup router, which I don't think anybody now
4947 uses, but I've fixed it anyway.
4949 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4950 well as to direct DNS lookups. Otherwise the handling of names in host
4951 lists is inconsistent and therefore confusing.
4957 PH/01 Two changes to the default runtime configuration:
4959 (1) Move the checks for relay_from_hosts and authenticated clients from
4960 after to before the (commented out) DNS black list checks.
4962 (2) Add control=submission to the relay_from_hosts and authenticated
4963 clients checks, on the grounds that messages accepted by these
4964 statements are most likely to be submissions.
4966 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4968 (1) Generate an error if the third argument for the ${prvs expansion is
4971 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4974 (3) Reset the variables that are obtained from the first argument of
4975 ${prvscheck and used in the second argument before leaving the code,
4976 because their memory is reclaimed, so using them afterwards may do
4979 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4980 one (it's much easier than Tom thought :-).
4982 (5) Because of (4), we can now allow for the use of $prvscheck_result
4983 inside the third argument.
4985 PH/03 For some reason, the default setting of PATH when running a command from
4986 a pipe transport was just "/usr/bin". I have changed it to
4989 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4990 anything to be listed in the output from -bV.
4992 PH/05 When a filter generated an autoreply, the entire To: header line was
4993 quoted in the delivery log line, like this:
4995 => >A.N.Other <ano@some.domain> <original@ddress> ...
4997 This has been changed so that it extracts the operative address. There
4998 may be more than one such address. If so, they are comma-separated, like
5001 => >ano@some.domain,ona@other.domain <original@ddress> ...
5003 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5004 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5005 not being looked up in the rDNS to get a host name, Exim was showing the
5006 IP address twice in Received: lines, even though the IP addresses were
5007 identical. For example:
5009 Received: from [1.2.3.4] (helo=[1.2.3.4])
5011 However, if the real host name was known, it was omitting the HELO data
5012 if it matched the actual IP address. This has been tidied up so that it
5013 doesn't show the same IP address twice.
5015 PH/07 When both +timestamp and +memory debugging was on, the value given by
5016 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5017 called by the debug printing, thereby overwriting the timestamp buffer.
5018 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5020 PH/08 When the original message was included in an autoreply transport, it
5021 always said "this is a copy of the message, including all the headers",
5022 even if body_only or headers_only was set. It now gives an appropriate
5025 PH/09 Applied a patch from the Sieve maintainer which:
5027 o fixes some comments
5028 o adds the (disabled) notify extension core
5029 o adds some debug output for the result of if/elsif tests
5030 o points to the current vacation draft in the documentation
5031 and documents the missing references header update
5035 o fixes a bug in processing the envelope test (when testing
5036 multiple envelope elements, the last element determined the
5039 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5040 Electronic Mail") by including:
5042 Auto-submitted: auto-generated
5044 in the messages that it generates (bounce messages and others, such as
5045 warnings). In the case of bounce messages for non-SMTP messages, there was
5046 also a typo: it was using "Auto_submitted" (underscore instead of
5047 hyphen). Since every message generated by Exim is necessarily in response
5048 to another message, thes have all been changed to:
5050 Auto-Submitted: auto-replied
5052 in accordance with these statements in the RFC:
5054 The auto-replied keyword:
5056 - SHOULD be used on messages sent in direct response to another
5057 message by an automatic process,
5059 - MUST NOT be used on manually-generated messages,
5061 - MAY be used on Delivery Status Notifications (DSNs) and Message
5062 Disposition Notifications (MDNs),
5064 - MUST NOT be used on messages generated by automatic or periodic
5065 processes, except for messages which are automatic responses to
5068 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5069 to the default Received: header definition.
5071 PH/12 Added log selector acl_warn_skipped (default on).
5073 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5074 variables because (a) they are in the wrong storage pool and (b) even if
5075 they were copied, it wouldn't work properly because of the caching.
5077 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5078 checking when decoding. Apparently there are clients that generate
5079 overlong encoded strings. Why am I not surprised?
5081 PH/15 If the first argument of "${if match_address" was not empty, but did not
5082 contain an "@" character, Exim crashed. Now it writes a panic log message
5083 and treats the condition as false.
5085 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5087 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5088 extension "envelope-auth". The code is finished and in agreement with
5089 other implementations, but there is no documentation so far and in fact,
5090 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5091 not changing the active code.
5093 Print executed "if" and "elsif" statements when debugging is used. This
5094 helps a great deal to understand what a filter does.
5096 Document more things not specified clearly in RFC3028. I had all this
5097 sorted out, when out of a sudden new issues came to my mind. Oops."
5099 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5102 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5103 canonical form (as documented). However, after a host name lookup from
5104 the IP address, check_host() was doing a simple string comparison with
5105 addresses acquired from the DNS when checking that the found name did
5106 have the original IP as one of its addresses. Since any found IPv6
5107 addresses are likely to be in abbreviated form, the comparison could
5108 fail. Luckily, there already exists a function for doing the comparison
5109 by converting both addresses to binary, so now that is used instead of
5110 the text comparison.
5112 PH/20 There was another similar case to PH/19, when a complete host name was
5113 given in a host list; looking up its IP address could give an abbreviated
5114 form, whereas the current host's name might or might not be abbreviated.
5115 The same fix has been applied.
5121 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5122 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5125 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5126 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5128 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5130 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5131 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5132 Also, for "Auto-Submitted", treat anything other than "no" as
5133 non-personal, in accordance with RFC 3834. (Previously it treated
5134 anything starting "auto-" as non-personal.)
5136 TF/01 The control=submission/name=... option had a problem with syntax
5137 errors if the name included a slash character. The /name= option
5138 now slurps the rest of the string, so it can include any characters
5139 but it must come last in the list of options (after /sender_retain
5142 PH/05 Some modifications to the interface to the fake nameserver for the new
5150 TK/01 Added the "success_on_redirect" address verification option. See
5151 NewStuff for rationale and an example.
5153 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5155 PH/02 Patch to exigrep to allow it to work on syslog lines.
5157 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5158 fread() to skip over the body file's header line, because in Cygwin the
5159 header line is locked and is inaccessible.
5161 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5162 co-exist for some time) to make it clear that it is the Exim ID that is
5163 referenced, not the Message-ID: header line.
5165 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5166 string_format() function, because snprintf() does not exist on all
5169 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5170 ${stat: expansion item.
5172 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5173 protocol synchronization error", to keep the pedants happy.
5175 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5176 well as for IRIX systems, when gcc is being used. See the host.c source
5179 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5181 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5184 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5185 transport and to the smtp transport in LMTP mode.
5187 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5189 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5190 run a filter in a subprocess. This could lead to confusion in subsequent
5191 lookups in the parent process. There should also be a search_tidyup() at
5192 the end of the subprocess.
5194 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5195 only if the host matched helo_try_verify_hosts, which caused the
5196 verification to occur when the EHLO/HELO command was issued. The ACL just
5197 tested the remembered result. Now, if a previous verification attempt has
5198 not happened, "verify = helo" does it there and then.
5200 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5202 TK/03 Fix log output including CR from clamd.
5204 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5205 exist provoked a memory error which could cause a segfault.
5207 PH/15 Installed PCRE 6.2
5209 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5211 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5212 of the problem. Specifically, suggested +O2 rather than +O1 for the
5215 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5217 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5218 record, the port was not correctly logged when the outgoing_port log
5219 selector was set (it logged the transort's default port).
5221 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5222 fallback_hosts, and "hosts" in the smtp transport.
5224 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5225 host errors such as "Connection refused".
5227 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5228 authentication with radiusclient 0.4.9:
5230 - Error returned from rc_read_config was caught wrongly
5231 - Username/password not passed on to radius server due to wrong length.
5233 The presumption is that some radiusclient API changes for 4.51/PH/17
5234 were not taken care of correctly. The code is still untested by me (my
5235 Linux distribution still has 0.3.2 of radiusclient), but it was
5236 contributed by a Radius user.
5238 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5239 expanding the "port" option of the smtp transport.
5241 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5242 while reading a MIME header. Thanks to Tom Hughes for a patch.
5244 PH/24 Include config.h inside local_scan.h so that configuration settings are
5247 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5248 that in an ACL for RCPT (for example), you can examine exactly what was
5251 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5252 commands, but it was not correctly comparing the address with the actual
5253 client host address. Thus, it would show the EHLO address in Received:
5254 header lines when this was not necessary.
5256 PH/27 Added the % operator to ${eval:}.
5258 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5259 it should be ignoring failures (because with -C, for example, it has lost
5260 privilege). It wasn't ignoring creation failures other than "already
5263 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5266 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5267 because an input line was too long, either on its own, or by virtue of
5268 too many continuations, the temporary file was not being removed, and the
5269 return code was incorrect.
5271 PH/31 Missing "BOOL" in function definition in filtertest.c.
5273 PH/32 Applied Sieve patches from the maintainer.
5275 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5277 PH/33 Added "verify = not_blind".
5279 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5280 Local/Makefile (with some defaults set). These are used in built scripts
5281 such as exicyclog, but they have never been used in the exim_install
5282 script (though there are many overriding facilities there). I have
5283 arranged that the exim_install script now takes note of these two
5286 PH/35 Installed configuration files for Dragonfly.
5288 PH/36 When a locally submitted message by a trusted user did not contain a
5289 From: header, and the sender address was obtained from -f or from an SMTP
5290 MAIL command, and the trusted user did not use -F to supply a sender
5291 name, $originator_name was incorrectly used when constructing a From:
5292 header. Furthermore, $originator_name was used for submission mode
5293 messages from external hosts without From: headers in a similar way,
5294 which is clearly wrong.
5296 PH/37 Added control=suppress_local_fixups.
5298 PH/38 When log_selector = +received_sender was set, and the addition of the
5299 sender made the log line's construction buffer exactly full, or one byte
5300 less than full, an overflow happened when the terminating "\n" was
5303 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5304 when the result of a list match is failure because a DNS lookup failed.
5306 PH/40 RM_COMMAND is now used in the building process.
5308 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5309 the "build-* directories that it finds.
5311 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5312 address in a domain literal was a prefix of an interface address.
5314 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5315 when verifying a sender address, unless rewrite_headers is false.
5317 PH/44 Wrote a long comment about why errors_to addresses are verified as
5318 recipients, not senders.
5320 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5321 the ratelimit ACL was added.
5323 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5325 PH/46 Added extra information about PostgreSQL errors to the error string.
5327 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5328 avoiding the need to install special zones in a real server. This is
5329 backwards compatible; if it can't find the fake resolver, it drops back.
5330 Thus, both old and new test suites can be run.
5332 TF/02 Added util/ratelimit.pl
5334 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5335 clock is set back in time.
5337 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5338 Candler <B.Candler@pobox.com>.
5340 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5341 OK for addresses that are the result of redirections.
5343 PH/48 A number of further additions for the benefit of the new test suite,
5344 including a fake gethostbyname() that interfaces to the fake DNS resolver
5347 TF/06 The fix for widen_domains has also been applied to qualify_single and
5348 search_parents which are the other dnslookup options that can cause
5351 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5354 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5356 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5357 service" if it encountered an MX record with an empty target hostname.
5358 The message is now "an MX or SRV record indicated no SMTP service".
5360 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5361 if the DNS of the sending site is misconfigured. This is quite a
5362 common situation. This change restores the behaviour of treating a
5363 helo verification defer as a failure.
5365 PH/51 If self=fail was set on a router, the bounce message did not include the
5366 actual error message.
5372 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5374 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5375 command itself ended in error, the underlying message about the transport
5376 filter timeout was being overwritten with the pipe command error. Now the
5377 underlying error message should be appended to the second error message.
5379 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5381 PH/02 Reduce the amount of output that "make" produces by default. Full output
5382 can still be requested.
5384 PH/03 The warning log line about a condition test deferring for a "warn" verb
5385 was being output only once per connection, rather than after each
5386 occurrence (because it was using the same function as for successful
5387 "warn" verbs). This seems wrong, so I have changed it.
5389 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5390 it should not have, which might have caused a crash in the right
5391 circumstances, but probably never did.
5393 PH/04 Installed a modified version of Tony Finch's patch to make submission
5394 mode fix the return path as well as the Sender: header line, and to
5395 add a /name= option so that you can make the user's friendly name appear
5398 TF/03 Added the control = fakedefer ACL modifier.
5400 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5401 Mark Lowes for thorough testing.
5403 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5405 TK/03 Merged latest SRS patch from Miles Wilton.
5407 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5408 with the definition in sysexits.h (which is #included earlier).
5409 Fortunately, Exim does not actually use EX_OK. The code used to try to
5410 preserve the sysexits.h value, by assuming that macro definitions were
5411 scanned for macro replacements. I have been disabused of this notion,
5412 so now the code just undefines EX_OK before #including unistd.h.
5414 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5415 in the smtp transport. When a block could not be written in a single
5416 write() function, the timeout was being re-applied to each part-write.
5417 This seems wrong - if the receiver was accepting one byte at a time it
5418 would take for ever. The timeout is now adjusted when this happens. It
5419 doesn't have to be particularly precise.
5421 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5422 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5424 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5425 without a domain if the check to <postmaster@domain> fails.
5427 SC/01 Eximstats: added -xls and the ability to specify output files
5428 (patch written by Frank Heydlauf).
5430 SC/02 Eximstats: use FileHandles for outputting results.
5432 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5434 SC/04 Eximstats: fixed display of large numbers with -nvr option
5436 SC/05 Eximstats: fixed merging of reports with empty tables.
5438 SC/06 Eximstats: added the -include_original_destination flag
5440 SC/07 Eximstats: removed tabs and trailing whitespace.
5442 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5444 TK/06 MBOX spool code: Add real "From " MBOX separator line
5445 so the .eml file is really in mbox format (even though
5446 most programs do not really care). Patch from Alex Miller.
5448 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5449 The latter is generated from $received_to and is only set if the
5450 message has one envelope recipient. SA can use these headers,
5451 obviously out-of-the-box. Patch from Alex Miller.
5453 PH/08 The ${def test on a variable was returning false if the variable's
5454 value was "0", contrary to what the specification has always said!
5455 The result should be true unless the variable is empty.
5457 PH/09 The syntax error of a character other than { following "${if
5458 def:variable_name" (after optional whitespace) was not being diagnosed.
5459 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5460 accidental colon was present, for example, could give incorrect results.
5462 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5463 result is used (not including appendfile, where other changes are about
5466 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5467 This involved changing a lot of size variables from int to off_t. It
5468 should work with maildirs and everything.
5470 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5471 spamd dying while we are connected to it.
5473 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5476 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5477 function for BDB 4.3.
5479 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5481 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5482 the sources for PCRE have been re-arranged and more files are now
5485 PH/15 The code I had for printing potentially long long variables in PH/11
5486 above was not the best (it lost precision). The length of off_t variables
5487 is now inspected at build time, and an appropriate printing format (%ld
5488 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5489 to be "long long int" or "long int". This is needed for the internal
5490 formatting function string_vformat().
5492 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5493 the configuration file to be ":syslog", then the script "guesses" where
5494 the logs files are, rather than using the compiled in default. In our
5495 case the guess is not the same as the compiled default, so the script
5496 suddenly stopped working when I started to use syslog. The patch checks
5497 to see if log_file_path is "". If so, it attempts to read it from exim
5498 with no configuration file to get the compiled in version, before it
5499 falls back to the previous guessing code."
5501 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5502 implementing BATV in an Exim configuration. See NewStuff for the gory
5505 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5506 Makefile that are specific to HP-UX.
5508 PH/18 If the "use_postmaster" option was set for a recipient callout together
5509 with the "random" option, the postmaster address was used as the MAIL
5510 FROM address for the random test, but not for the subsequent recipient
5511 test. It is now used for both.
5513 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5514 patch removes a few documentation additions to RFC 3028, because the
5515 latest draft now contains them. It adds the new en;ascii-case comparator
5516 and a new error check for 8bit text in MIME parts. Comparator and
5517 require names are now matched exactly. I enabled the subaddress
5518 extension, but it is not well tested yet (read: it works for me)."
5520 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5521 rework some of the code of TK/09 above to avoid the hardwired use of
5522 "%lld" and "long long". Replaced the call to snprintf() with a call to
5525 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5526 records point to non-existent hosts", "retry timeout exceeded", and
5527 "retry time not reached for any host after a long failure period".
5529 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5530 experimental DomainKeys support:
5532 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5533 (2) On an error such as an illegally used "control", the wrong name for
5534 the control was given.
5536 These problems did NOT occur unless DomainKeys support was compiled.
5538 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5540 PH/24 Added ${if match_ip condition.
5542 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5543 databases so that it will be absolutely obvious if a crash occurs in the
5544 DB library. This is a regular occurrence (often caused by mis-matched
5547 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5548 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5549 picked up on a user's system that detects such things. There doesn't seem
5550 to be a gcc warning option for this - only an attribute that has to be
5551 put on the function's prototype. It seems that in Fedora Core 4 they have
5552 set this on a number of new functions. No doubt there will be more in due
5555 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5556 specify a transport. However, if an address that was verified by such a
5557 router was the subject of a callout, Exim crashed because it tried to
5558 read the rcpt_include_affixes from the non-existent transport. Now it
5559 just assumes that the setting of that option is false. This bug was
5560 introduced by 4.51/PH/31.
5562 PH/28 Changed -d+all to exclude +memory, because that information is very
5563 rarely of interest, but it makes the output a lot bigger. People tend to
5564 do -d+all out of habit.
5566 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5567 code in os-type was giving problems when libc.so lives in lib64, like on
5570 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5571 aren't the modern standard, and it seems that some systems' include files
5572 don't always have them. Exim was already checking for some of the newer
5573 ones like T_AAAA, and defining it itself. I've added checks for all the
5574 record types that Exim uses.
5576 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5577 not automatically generating a new one, as it is supposed to. This
5578 prevented TLS from working. If the file did exist, but contained invalid
5579 data, a new version was generated, as expected. It was only the case of a
5580 non-existent file that was broken.
5582 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5583 with a change in libdomainkeys > 0.64.
5585 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5586 from DNS. If the selector record carries the flag, it now has
5587 precedence over the domain-wide flag.
5589 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5591 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5592 the use of a port name that isn't defined in /etc/services) occurred, the
5593 message was deferred as in a normal delivery, and thus remained on the
5594 spool, instead of being failed because of the mua_wrapper setting. This
5595 is now fixed, and I tidied up some of the mua_wrapper messages at the
5598 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5599 the messages in a hash of arrays rather than using individual hashes.
5600 This is a bit cleaner and results in dramatic memory savings, albeit
5601 at a slight CPU cost.
5603 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5604 as requested by Marc Sherman.
5606 SC/10 Eximstats: added histograms for user specified patterns as requested
5609 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5611 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5612 fopen() in the content-scanning modules that did not already have it.
5618 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5619 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5621 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5623 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5625 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5626 to test Sieve filters that use "vacation".
5628 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5629 that changes the way the GnuTLS parameters are stored in the cache file.
5630 The new format can be generated externally. For backward compatibility,
5631 if the data in the cache doesn't make sense, Exim assumes it has read an
5632 old-format file, and it generates new data and writes a new file. This
5633 means that you can't go back to an older release without removing the
5636 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5637 work if there are any delivery delays because "one_time" forces the
5638 parent to be marked "delivered", so its unseen clone is never tried
5639 again. For this reason, Exim now forbids the simultaneous setting of
5642 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5643 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5644 entry if you want to know the details. The fix, however, neglected to
5645 consider the case when local delivery batching is involved. The test for
5646 "previously delivered" was not happening when checking to see if an
5647 address could be batched with a previous (undelivered) one; under
5648 certain circumstances this could lead to multiple deliveries to the same
5651 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5652 in its include files, and this causes problems building Exim.
5654 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5655 header_syntax) but Exim was just ignoring anything given after a slash.
5656 In particular, this caused confusion with an attempt to use "verify =
5657 reverse_host_lookup/defer_ok". An error is now given when options are
5658 supplied for verify items that do not have them. (Maybe reverse_host_
5659 lookup should have a defer_ok option, but that's a different point.)
5661 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5662 defined by RFC 821) to 2048, because there were problems with some AUTH
5663 commands, and RFC 1869 says the size should be increased for extended
5664 SMTP commands that take arguments.
5666 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5669 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5670 "unknown" error; now it says that the functionality isn't in the binary.
5672 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5673 an address' error message when a string expansion fails (syntax or
5674 whatever). Otherwise the password may appear in the log. Following change
5675 PH/42 below, there is no longer a chance of it appearing in a bounce
5678 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5680 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5681 of its addresses was ever tried. (Bugzilla bug #2.)
5683 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5684 the result incorrectly in the debug output. (It correctly added a newline
5685 to what was transported.)
5687 TF/01 Added $received_time.
5689 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5690 commented out examples of how to interface to a virus scanner and to
5691 SpamAssassin. Also added commented examples of av_scanner and
5692 spamd_address settings.
5694 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5695 and controls are allowed in which ACLs. There were a couple of minor
5696 errors. Some of the entries in the conditions table (which is a table of
5697 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5698 negation of where the condition IS allowed.
5700 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5702 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5703 header file does not have a version number, so I've had to invent a new
5704 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5705 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5706 radiusclient), but it was contributed by a Radius user.
5708 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5709 files or directories, for OpenSSL.
5711 PH/19 When an Exim process that is running as root has to create an Exim log
5712 file, it does so in a subprocess that runs as exim:exim so as to get the
5713 ownership right at creation (otherwise, other Exim processes might see
5714 the file with the wrong ownership). There was no test for failure of this
5715 fork() call, which would lead to the process getting stuck as it waited
5716 for a non-existent subprocess. Forks do occasionally fail when resources
5717 run out. I reviewed all the other calls to fork(); they all seem to check
5720 PH/20 When checking for unexpected SMTP input at connect time (before writing
5721 the banner), Exim was not dealing correctly with a non-positive return
5722 from the read() function. If the client had disconnected by this time,
5723 the result was a log entry for a synchronization error with an empty
5724 string after "input=" when read() returned zero. If read() returned -1
5725 (an event I could not check), uninitialized data bytes were printed.
5726 There were reports of junk text (parts of files, etc) appearing after
5729 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5731 PH/22 Added support for macro redefinition, and (re)definition in between
5732 driver and ACL definitions.
5734 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5735 forgetting to use the resulting value; it was using the unexpanded value.
5737 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5738 hadn't been configured. The fix is from Juergen Kreileder, who
5739 understands it better than I do:
5741 "Here's what I see happening with three configured cyrus_sasl
5742 authenticators configured (plain, login, cram-md5):
5744 On startup auth_cyrus_sasl_init() gets called for each of these.
5745 This means three calls to sasl_listmech() without a specified mech_list.
5746 => SASL tests which mechs of all available mechs actually work
5747 => three warnings about OTP not working
5748 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5750 With the patch, sasl_listmech() also gets called three times. But now
5751 SASL's mech_list option is set to the server_mech specified in the the
5752 authenticator. Or in other words, the answer from sasl_listmech()
5753 gets limited to just the mech you're testing for (which is different
5755 => the return list contains just 'plain' or 'login', 'cram-md5' or
5756 nothing depending on the value of ob->server_mech.
5758 I've just tested the patch: Authentication still works fine,
5759 unavailable mechs specified in the exim configuration are still
5760 caught, and the auth.log warnings about OTP are gone."
5762 PH/25 When debugging is enabled, the contents of the command line are added
5763 to the debugging output, even when log_selector=+arguments is not
5766 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5767 answer is "GNU", and only if the return is "GNU/something" is the answer
5770 PH/27 $acl_verify_message is now set immediately after the failure of a
5771 verification in an ACL, and so is available in subsequent modifiers. In
5772 particular, the message can be preserved by coding like this:
5774 warn !verify = sender
5775 set acl_m0 = $acl_verify_message
5777 Previously, $acl_verify_message was set only while expanding "message"
5778 and "log_message" when a very denied access.
5780 PH/28 Modified OS/os.c-Linux with
5782 -#ifndef OS_LOAD_AVERAGE
5783 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5785 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5786 nomenclature these days.)
5788 PH/29 Installed patch from the Sieve maintainer that adds the options
5789 sieve_useraddress and sieve_subaddress to the redirect router.
5791 PH/30 In these circumstances:
5792 . Two addresses routed to the same list of hosts;
5793 . First host does not offer TLS;
5794 . First host accepts first address;
5795 . First host gives temporary error to second address;
5796 . Second host offers TLS and a TLS session is established;
5797 . Second host accepts second address.
5798 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5799 and peerdn, if requested) that were in fact used only for the second
5802 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5803 attention to any local part prefix or suffix that was matched by the
5804 router that accepted the address. It now behaves in the same way as it
5805 does for delivery: the affixes are removed from the local part unless
5806 rcpt_include_affixes is set on the transport.
5808 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5809 timeout during the DATA phase of an incoming message.
5811 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5812 applied a patch sanctioned by the Sieve maintainer.
5814 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5815 the uid or gid is negative. A case of a negative gid caused this to be
5816 noticed. The fix allows for either to be negative.
5818 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5819 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5822 PH/36 The change PH/12 above was broken. Fixed it.
5824 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5825 the grounds that routing the same address twice would always produce the
5826 same answer. This might have been true once, but it is certainly no
5827 longer true now. Routing a child address may depend on the previous
5828 routing that produced that child. Some complicated redirection strategies
5829 went wrong when messages had multiple recipients, and made Exim's
5830 behaviour dependent on the order in which the addresses were given.
5832 I have moved the duplicate checking until after the routing is complete.
5833 Exim scans the addresses that are assigned to local and remote
5834 transports, and removes any duplicates. This means that more work will be
5835 done, as duplicates will always all be routed, but duplicates are
5836 presumably rare, so I don't expect this is of any significance.
5838 For deliveries to pipes, files, and autoreplies, the duplicate checking
5839 still happens during the routing process, since they are not going to be
5842 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5843 It corrects a timeout issue with spamd. This is Ian's comment: "The
5844 background is that sometimes spamd either never reads data from a
5845 connection it has accepted, or it never writes response data. The exiscan
5846 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5847 blindly assumes that writes won't block so it may never time out."
5849 PH/39 Allow G after quota size as well as K and M.
5851 PH/40 The value set for $authenticated_id in an authenticator may not contain
5852 binary zeroes or newlines because the value is written to log lines and
5853 to spool files. There was no check on this. Now the value is run through
5854 the string_printing() function so that such characters are converted to
5855 printable escape sequences.
5857 PH/41 $message_linecount is a new variable that contains the total number of
5858 lines in the message. Compare $body_linecount, which is the count for the
5861 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5862 bounce and delay warning messages, except in certain special cases, which
5865 (a) An SMTP error message from a remote host;
5866 (b) A message specified in a :fail: redirection;
5867 (c) A message specified in a "fail" command in a system filter;
5868 (d) A message specified in a FAIL return from the queryprogram router;
5869 (e) A message specified by the cannot_route_message router option.
5871 In these cases only, Exim does include the error details in bounce and
5872 warning messages. There are also a few cases where bland messages such
5873 as "unrouteable address" or "local delivery error" are given.
5875 PH/43 $value is now also set for the "else" part of a ${run expansion.
5877 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5878 being worked on, but at least Exim now implements the latest version to
5881 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5882 process to complete was treated as a delivery failure, a timeout while
5883 writing the message to the pipe was logged, but erroneously treated as a
5884 successful delivery. Such timeouts include transport filter timeouts. For
5885 consistency with the overall process timeout, these timeouts are now
5886 treated as errors, giving rise to delivery failures by default. However,
5887 there is now a new Boolean option for the pipe transport called
5888 timeout_defer, which, if set TRUE, converts the failures into defers for
5889 both kinds of timeout. A transport filter timeout is now identified in
5892 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5893 systems where "make" and "gmake" are different, calling "gmake" at top
5894 level broke things. I've arranged for the value of $(MAKE) to be passed
5895 from the Makefile to this script so that it can call the same version of
5899 A note about Exim versions 4.44 and 4.50
5900 ----------------------------------------
5902 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5903 changes of various kinds. As a consequence, a big documentation update was
5904 needed. This delayed the release for rather longer than seemed good, especially
5905 in the light of a couple of (minor) security issues. Therefore, the changes
5906 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5907 release. So 4.44 and 4.50 are in effect two different branches that both start
5910 I have left the 4.50 change log unchanged; it contains all the changes since
5911 4.43. The change log for 4.44 is below; many of its items are identical to
5912 those for 4.50. This seems to be the most sensible way to preserve the
5913 historical information.
5919 1. Minor wording change to the doc/README.SIEVE file.
5921 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5922 computation of the current number of files was incorrect.
5924 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5925 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5928 4. Give more explanation in the error message when the command for a transport
5929 filter fails to execute.
5931 5. There are several places where Exim runs a non-Exim command in a
5932 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5933 was being done only for the command run by the queryprogram router. It is
5934 now done for all such subprocesses. The other cases are: ${run, transport
5935 filters, and the commands run by the lmtp and pipe transports.
5937 6. Added CONFIGURE_GROUP build-time option.
5939 7. Some older OS have a limit of 256 on the maximum number of file
5940 descriptors. Exim was using setrlimit() to set 1000 as a large value
5941 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5942 systems. I've change it so that if it can't get 1000, it tries for 256.
5944 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5945 was an oversight, and furthermore, ever since the addition of extra
5946 controls (e.g. 4.43/32), the checks on when to allow different forms of
5947 "control" were broken. There should now be diagnostics for all cases when a
5948 control that does not make sense is encountered.
5950 9. Added the /retain_sender option to "control=submission".
5952 10. $recipients is now available in the predata ACL (oversight).
5954 11. Tidy the search cache before the fork to do a delivery from a message
5955 received from the command line. Otherwise the child will trigger a lookup
5956 failure and thereby defer the delivery if it tries to use (for example) a
5957 cached ldap connection that the parent has called unbind on.
5959 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5960 of $address_data from the recipient verification was clobbered by the
5961 sender verification.
5963 13. The value of address_data from a sender verification is now available in
5964 $sender_address_data in subsequent conditions in the ACL statement.
5966 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5968 15. Added a new option "connect=<time>" to callout options, to set a different
5971 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5972 was its contents. (It was OK if the option was not defined at all.)
5974 17. A "Completed" log line is now written for messages that are removed from
5975 the spool by the -Mrm option.
5977 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5978 information about exactly what failed.
5980 19. Added -dd to debug only the daemon process.
5982 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5983 handles timeouts, both on the server side and network timeouts. Renamed the
5984 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5986 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5987 It is now set to "smtps".
5989 22. $host_address is now set to the target address during the checking of
5990 ignore_target_hosts.
5992 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5993 being passed; this would have caused $sender_host_name to have been used if
5994 matching the list had actually called for a host name (not very likely,
5995 since this list is usually IP addresses). A host name is now passed as
5998 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5999 code that specifies a non-restarting handler (typically sigaction() in
6000 modern systems) in an attempt to fix a rare and obscure crash bug.
6002 25. Narrowed the window for a race in the daemon that could cause it to ignore
6003 SIGCHLD signals. This is not a major problem, because they are used only to
6004 wake it up if nothing else does.
6006 26. A malformed maildirsize file could cause Exim to calculate negative values
6007 for the mailbox size or file count. Odd effects could occur as a result.
6008 The maildirsize information is now recalculated if the size or filecount
6011 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6012 support for a long time. Removed HAVE_SYS_VFS_H.
6014 28. Installed the latest version of exipick from John Jetmore.
6016 29. In an address list, if the pattern was not a regular expression, an empty
6017 subject address (from a bounce message) matched only if the pattern was an
6018 empty string. Non-empty patterns were not even tested. This was the wrong
6019 because it is perfectly reasonable to use an empty address as part of a
6020 database query. An empty address is now tested by patterns that are
6021 lookups. However, all the other forms of pattern expect the subject to
6022 contain a local part and a domain, and therefore, for them, an empty
6023 address still always fails if the pattern is not itself empty.
6025 30. Exim went into a mad DNS loop when attempting to do a callout where the
6026 host was specified on an smtp transport, and looking it up yielded more
6027 than one IP address.
6029 31. Re-factored the code for checking spool and log partition space into a
6030 function that finds that data and another that does the check. The former
6031 is then used to implement four new variables: $spool_space, $log_space,
6032 $spool_inodes, and $log_inodes.
6034 32. The RFC2047 encoding function was originally intended for short strings
6035 such as real names; it was not keeping to the 75-character limit for
6036 encoded words that the RFC imposes. It now respects the limit, and
6037 generates multiple encoded words if necessary. To be on the safe side, I
6038 have increased the buffer size for the ${rfc2047: expansion operator from
6041 33. It is now permitted to omit both strings after an "if" condition; if the
6042 condition is true, the result is "true". As before, when the second string
6043 is omitted, a false condition yields an empty string. This makes it less
6044 cumbersome to write custom ACL and router conditions.
6046 34. Failure to deliver a bounce message always caused it to be frozen, even if
6047 there was an errors_to setting on the router. The errors_to setting is now
6050 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6051 canonical form (fully expanded) before being placed in
6052 $sender_host_address.
6054 36. The table in the code that translates DNS record types into text (T_A to
6055 "A" for instance) was missing entries for NS and CNAME. It is just possible
6056 that this could have caused confusion if both these types were looked up
6057 for the same domain, because the text type is used as part of Exim's
6058 per-process caching. But the chance of anyone hitting this buglet seems
6061 37. The dnsdb lookup has been extended in a number of ways.
6063 (1) There is a new type, "zns", which walks up the domain tree until it
6064 finds some nameserver records. It should be used with care.
6066 (2) There is a new type, "mxh", which is like "mx" except that it returns
6067 just the host names, not the priorities.
6069 (3) It is now possible to give a list of domains (or IP addresses) to be
6070 looked up. The behaviour when one of the lookups defers can be
6071 controlled by a keyword.
6073 (4) It is now possible to specify the separator character for use when
6074 multiple records are returned.
6076 38. The dnslists ACL condition has been extended: it is now possible to supply
6077 a list of IP addresses and/or domains to be looked up in a particular DNS
6080 39. Added log_selector=+queue_time_overall.
6082 40. When running the queue in the test harness, wait just a tad after forking a
6083 delivery process, to get repeatability of debugging output.
6085 41. Include certificate and key file names in error message when GnuTLS fails
6086 to set them up, because the GnuTLS error message doesn't include the name
6087 of the failing file when there is a problem reading it.
6089 42. Allow both -bf and -bF in the same test run.
6091 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6093 44. The "Exiscan patch" is now merged into the mainline Exim source.
6095 45. Sometimes the final signoff response after QUIT could fail to get
6096 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6097 < 0 before doing a fflush(). This bug looks as though it goes back to the
6098 introduction of TLS in release 3.20, but "sometimes" must have been rare
6099 because the tests only now provoked it.
6101 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6102 (this can affect the format of dates).
6104 47. exim_tidydb, when checking for the continued existence of a message for
6105 which it has found a message-specific retry record, was not finding
6106 messages that were in split spool directories. Consequently, it was
6107 deleting retry records that should have stayed in existence.
6109 48. Steve fixed some bugs in eximstats.
6111 49. The SPA authentication driver was not abandoning authentication and moving
6112 on to the next authenticator when an expansion was forced to fail,
6113 contradicting the general specification for all authenticators. Instead it
6114 was generating a temporary error. It now behaves as specified.
6116 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6117 (the order specifies the preference for clients). The order is now AES256,
6118 AES128, 3DES, ARCFOUR128.
6120 51. Small patch to Sieve code - explicitly set From: when generating an
6123 52. Exim crashed if a remote delivery caused a very long error message to be
6124 recorded - for instance if somebody sent an entire SpamAssassin report back
6125 as a large number of 550 error lines. This bug was coincidentally fixed by
6126 increasing the size of one of Exim's internal buffers (big_buffer) that
6127 happened as part of the Exiscan merge. However, to be on the safe side, I
6128 have made the code more robust (and fixed the comments that describe what
6131 53. Now that there can be additional text after "Completed" in log lines (if
6132 the queue_time_overall log selector is set), a one-byte patch to exigrep
6133 was needed to allow it to recognize "Completed" as not the last thing in
6136 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6137 patch that reportedly fixes this has been added. I am not expert enough to
6138 create a test for it. This is what the patch creator wrote:
6140 "I found a little strange behaviour of ldap code when working with
6141 Windows 2003 AD Domain, where users was placed in more than one
6142 Organization Units. When I tried to give exim partial DN, the exit code
6143 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6144 But simultaneously result of request was absolutely normal ldap result,
6145 so I produce this patch..."
6147 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6148 REFERENCE, so I have modified the code to exclude the patch when that macro
6151 55. Some experimental protocols are using DNS PTR records for new purposes. The
6152 keys for these records are domain names, not reversed IP addresses. The
6153 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6154 leaves it alone. Component reversal etc. now happens only for IP addresses.
6157 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6159 57. Double the size of the debug message buffer (to 2048) so that more of very
6160 long debug lines gets shown.
6162 58. The exicyclog utility now does better if the number of log files to keep
6163 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6165 59. Two changes related to the smtp_active_hostname option:
6167 (1) $smtp_active_hostname is now available as a variable.
6168 (2) The default for smtp_banner uses $smtp_active_hostname instead
6169 of $primary_hostname.
6171 60. The host_aton() function is supposed to be passed a string that is known
6172 to be a valid IP address. However, in the case of IPv6 addresses, it was
6173 not checking this. This is a hostage to fortune. Exim now panics and dies
6174 if the condition is not met. A case was found where this could be provoked
6175 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6176 components; fortuitously, this particular loophole had already been fixed
6177 by change 4.50/55 above.
6179 If there are any other similar loopholes, the new check in host_aton()
6180 itself should stop them being exploited. The report I received stated that
6181 data on the command line could provoke the exploit when Exim was running as
6182 exim, but did not say which command line option was involved. All I could
6183 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6184 running as the user.
6187 61. There was a buffer overflow vulnerability in the SPA authentication code
6188 (which came originally from the Samba project). I have added a test to the
6189 spa_base64_to_bits() function which I hope fixes it.
6192 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6193 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6195 63. The daemon start-up calls getloadavg() while still root for those OS that
6196 need the first call to be done as root, but it missed one case: when
6197 deliver_queue_load_max is set with deliver_drop_privilege. This is
6198 necessary for the benefit of the queue runner, because there is no re-exec
6199 when deliver_drop_privilege is set.
6201 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6202 This has been fixed.
6204 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6205 was in use, was not putting the data itself into the right store pool;
6206 consequently, it could be overwritten for a subsequent message in the same
6207 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6210 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6212 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6213 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6214 the function were treating the return as a boolean value, which happened to
6215 work because 0=false and not-0=true, but is not correct code.
6217 68. The host_aton() function was not handling scoped IPv6 addresses (those
6218 with, for example, "%eth0" on the end) correctly.
6220 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6221 negated items (that is, ~something) in unsigned ints. Some compilers
6222 apparently mutter when there is no cast.
6224 70. If an address verification called from an ACL failed, and did not produce a
6225 user-specific message (i.e. there was only a "system" message), nothing was
6226 put in $acl_verify_message. In this situation, it now puts the system
6229 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6230 session; change 4.31/43 added the unwanted input to the log line - except
6231 that it did not do this in the start of session case. It now does.
6233 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6234 This is wrong and can cause the other end to generate a synchronization
6235 error if it is another Exim or anything else that does the synchronization
6236 check. A QUIT command is no longer sent after a timeout.
6238 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6239 during host lookups.
6241 74. The defer_ok option of callout verification was not working if it was used
6242 when verifying addresses in header lines, that is, for this case:
6244 verify = header_sender/callout=defer_ok
6246 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6247 those file descriptors could be used for SMTP connections. If anything
6248 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6249 could be sent to the SMTP client, causing chaos. The daemon now opens
6250 stdin, stdout, and stderr to /dev/null when it puts itself into the
6253 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6254 log by default. The user can override this with suitable Perl magic.
6256 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6257 the log message when discard triggers, was not working for the DATA ACL or
6258 for the non-SMTP ACL.
6260 78. Error message wording change in sieve.c.
6262 79. If smtp_accept_max_per_host was set, the number of connections could be
6263 restricted to fewer than expected, because the daemon was trying to set up
6264 a new connection before checking whether the processes handling previous
6265 connections had finished. The check for completed processes is now done
6266 earlier. On busy systems, this bug wouldn't be noticed because something
6267 else would have woken the daemon, and it would have reaped the completed
6270 80. If a message was submitted locally by a user whose login name contained one
6271 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6272 It caused a spool format error. I have fixed the spool reading code. A
6273 related problem was that the "from" clause in the Received: line became
6274 illegal because of the space(s). It is now covered by ${quote_local_part.
6276 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6279 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6281 83. Patch from Sieve maintainer to fix unterminated string problem in
6282 "vacation" handling.
6284 84. Some minor changes to the Linux configuration files to help with other
6285 OS variants using glibc.
6287 85. One more patch for Sieve to update vacation handling to latest spec.
6290 ----------------------------------------------------
6291 See the note above about the 4.44 and 4.50 releases.
6292 ----------------------------------------------------
6298 1. Change 4.43/35 introduced a bug that caused file counts to be
6299 incorrectly computed when quota_filecount was set in an appendfile
6302 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6303 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6306 3. Give more explanation in the error message when the command for a transport
6307 filter fails to execute.
6309 4. There are several places where Exim runs a non-Exim command in a
6310 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6311 was being done only for the command run by the queryprogram router. It is
6312 now done for all such subprocesses. The other cases are: ${run, transport
6313 filters, and the commands run by the lmtp and pipe transports.
6315 5. Some older OS have a limit of 256 on the maximum number of file
6316 descriptors. Exim was using setrlimit() to set 1000 as a large value
6317 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6318 systems. I've change it so that if it can't get 1000, it tries for 256.
6320 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6321 was an oversight, and furthermore, ever since the addition of extra
6322 controls (e.g. 4.43/32), the checks on when to allow different forms of
6323 "control" were broken. There should now be diagnostics for all cases when a
6324 control that does not make sense is encountered.
6326 7. $recipients is now available in the predata ACL (oversight).
6328 8. Tidy the search cache before the fork to do a delivery from a message
6329 received from the command line. Otherwise the child will trigger a lookup
6330 failure and thereby defer the delivery if it tries to use (for example) a
6331 cached ldap connection that the parent has called unbind on.
6333 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6334 of $address_data from the recipient verification was clobbered by the
6335 sender verification.
6337 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6338 was its contents. (It was OK if the option was not defined at all.)
6340 11. A "Completed" log line is now written for messages that are removed from
6341 the spool by the -Mrm option.
6343 12. $host_address is now set to the target address during the checking of
6344 ignore_target_hosts.
6346 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6347 being passed; this would have caused $sender_host_name to have been used if
6348 matching the list had actually called for a host name (not very likely,
6349 since this list is usually IP addresses). A host name is now passed as
6352 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6353 code that specifies a non-restarting handler (typically sigaction() in
6354 modern systems) in an attempt to fix a rare and obscure crash bug.
6356 15. Narrowed the window for a race in the daemon that could cause it to ignore
6357 SIGCHLD signals. This is not a major problem, because they are used only to
6358 wake it up if nothing else does.
6360 16. A malformed maildirsize file could cause Exim to calculate negative values
6361 for the mailbox size or file count. Odd effects could occur as a result.
6362 The maildirsize information is now recalculated if the size or filecount
6365 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6366 support for a long time. Removed HAVE_SYS_VFS_H.
6368 18. Updated exipick to current release from John Jetmore.
6370 19. Allow an empty sender to be matched against a lookup in an address list.
6371 Previously the only cases considered were a regular expression, or an
6374 20. Exim went into a mad DNS lookup loop when doing a callout where the
6375 host was specified on the transport, if the DNS lookup yielded more than
6378 21. The RFC2047 encoding function was originally intended for short strings
6379 such as real names; it was not keeping to the 75-character limit for
6380 encoded words that the RFC imposes. It now respects the limit, and
6381 generates multiple encoded words if necessary. To be on the safe side, I
6382 have increased the buffer size for the ${rfc2047: expansion operator from
6385 22. Failure to deliver a bounce message always caused it to be frozen, even if
6386 there was an errors_to setting on the router. The errors_to setting is now
6389 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6390 canonical form (fully expanded) before being placed in
6391 $sender_host_address.
6393 24. Updated eximstats to version 1.33
6395 25. Include certificate and key file names in error message when GnuTLS fails
6396 to set them up, because the GnuTLS error message doesn't include the name
6397 of the failing file when there is a problem reading it.
6399 26. Expand error message when OpenSSL has problems setting up cert/key files.
6402 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6403 (this can affect the format of dates).
6405 28. exim_tidydb, when checking for the continued existence of a message for
6406 which it has found a message-specific retry record, was not finding
6407 messages that were in split spool directories. Consequently, it was
6408 deleting retry records that should have stayed in existence.
6410 29. eximstats updated to version 1.35
6411 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6412 1.35 - bugfix such that pie charts by volume are generated correctly
6414 30. The SPA authentication driver was not abandoning authentication and moving
6415 on to the next authenticator when an expansion was forced to fail,
6416 contradicting the general specification for all authenticators. Instead it
6417 was generating a temporary error. It now behaves as specified.
6419 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6420 (the order specifies the preference for clients). The order is now AES256,
6421 AES128, 3DES, ARCFOUR128.
6423 31. Small patch to Sieve code - explicitly set From: when generating an
6426 32. Exim crashed if a remote delivery caused a very long error message to be
6427 recorded - for instance if somebody sent an entire SpamAssassin report back
6428 as a large number of 550 error lines. This bug was coincidentally fixed by
6429 increasing the size of one of Exim's internal buffers (big_buffer) that
6430 happened as part of the Exiscan merge. However, to be on the safe side, I
6431 have made the code more robust (and fixed the comments that describe what
6434 33. Some experimental protocols are using DNS PTR records for new purposes. The
6435 keys for these records are domain names, not reversed IP addresses. The
6436 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6437 leaves it alone. Component reversal etc. now happens only for IP addresses.
6440 34. The host_aton() function is supposed to be passed a string that is known
6441 to be a valid IP address. However, in the case of IPv6 addresses, it was
6442 not checking this. This is a hostage to fortune. Exim now panics and dies
6443 if the condition is not met. A case was found where this could be provoked
6444 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6445 components; fortuitously, this particular loophole had already been fixed
6446 by change 4.50/55 or 4.44/33 above.
6448 If there are any other similar loopholes, the new check in host_aton()
6449 itself should stop them being exploited. The report I received stated that
6450 data on the command line could provoke the exploit when Exim was running as
6451 exim, but did not say which command line option was involved. All I could
6452 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6453 running as the user.
6456 35. There was a buffer overflow vulnerability in the SPA authentication code
6457 (which came originally from the Samba project). I have added a test to the
6458 spa_base64_to_bits() function which I hope fixes it.
6461 36. The daemon start-up calls getloadavg() while still root for those OS that
6462 need the first call to be done as root, but it missed one case: when
6463 deliver_queue_load_max is set with deliver_drop_privilege. This is
6464 necessary for the benefit of the queue runner, because there is no re-exec
6465 when deliver_drop_privilege is set.
6467 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6468 was in use, was not putting the data itself into the right store pool;
6469 consequently, it could be overwritten for a subsequent message in the same
6470 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6473 38. Sometimes the final signoff response after QUIT could fail to get
6474 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6475 < 0 before doing a fflush(). This bug looks as though it goes back to the
6476 introduction of TLS in release 3.20, but "sometimes" must have been rare
6477 because the tests only now provoked it.
6483 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6484 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6485 Now it returns an integer. A number of calls were still expecting a T/F
6486 return. Fortuitously, in all cases, the tests worked in OK situations,
6487 which is the norm. However, things would have gone wrong on any write
6488 failures on the smtp file descriptor. This function is used when sending
6489 messages over SMTP and also when doing verify callouts.
6491 2. When Exim is called to do synchronous delivery of a locally submitted
6492 message (the -odf or -odi options), it no longer closes stderr before doing
6495 3. Implemented the mua_wrapper option.
6497 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6499 5. Implemented the functions header_remove(), header_testname(),
6500 header_add_at_position(), and receive_remove_recipient(), and exported them
6503 6. If an ACL "warn" statement specified the addition of headers, Exim already
6504 inserted X-ACL-Warn: at the start if there was no header name. However, it
6505 was not making this test for the second and subsequent header lines if
6506 there were newlines in the string. This meant that an invalid header could
6507 be inserted if Exim was badly configured.
6509 7. Allow an ACL "warn" statement to add header lines at the start or after all
6510 the Received: headers, as well as at the end.
6512 8. Added the rcpt_4xx retry error code.
6514 9. Added postmaster_mailfrom=xxx to callout verification option.
6516 10. Added mailfrom=xxxx to the callout verification option, for verify=
6519 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6520 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6522 12. Inserted some casts to stop certain compilers warning when using pointer
6523 differences as field lengths or precisions in printf-type calls (mostly
6524 affecting debugging statements).
6526 13. Added optional readline() support for -be (dynamically loaded).
6528 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6529 same clock tick as a message's arrival, so that its received time was the
6530 same as the "first fail" time on the retry record, and that message
6531 remained on the queue past the ultimate address timeout, every queue runner
6532 would try a delivery (because it was past the ultimate address timeout) but
6533 after another failure, the ultimate address timeout, which should have then
6534 bounced the address, did not kick in. This was a "< instead of <=" error;
6535 in most cases the first failure would have been in the next clock tick
6536 after the received time, and all would be well.
6538 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6539 being recognized when the domain list was tested by the match_domain
6540 condition in an expansion string.
6542 16. Added the ${str2b64: operator.
6544 17. Exim was always calling setrlimit() to set a large limit for the number of
6545 processes, without checking whether the existing limit was already
6546 adequate. (It did check for the limit on file descriptors.) Furthermore,
6547 errors from getrlimit() and setrlimit() were being ignored. Now they are
6548 logged to the main and panic logs, but Exim does carry on, to try to do its
6549 job under whatever limits there are.
6551 18. Imported PCRE 5.0.
6553 19. Trivial typo in log message " temporarily refused connection" (the leading
6556 20. If the log selector return_path_on_delivery was set and an address was
6557 redirected to /dev/null, the delivery process crashed because it assumed
6558 that a return path would always be set for a "successful" delivery. In this
6559 case, the whole delivery is bypassed as an optimization, and therefore no
6562 21. Internal re-arrangement: the function for sending a challenge and reading
6563 a response while authentication was assuming a zero-terminated challenge
6564 string. It's now changed to take a pointer and a length, to allow for
6565 binary data in such strings.
6567 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6569 23. Exim was not respecting finduser_retries when seeking the login of the
6570 uid under which it was called; it was always trying 10 times. (The default
6571 setting of finduser_retries is zero.) Also, it was sleeping after the final
6572 failure, which is pointless.
6574 24. Implemented tls_on_connect_ports.
6576 25. Implemented acl_smtp_predata.
6578 26. If the domain in control=submission is set empty, Exim assumes that the
6579 authenticated id is a complete email address when it generates From: or
6580 Sender: header lines.
6582 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6583 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6584 chown and chgrp in /bin and hostname in /usr/bin.
6586 28. Exim was keeping the "process log" file open after each use, just as it
6587 does for the main log. This opens the possibility of it remaining open for
6588 long periods when the USR1 signal hits a daemon. Occasional processlog
6589 errors were reported, that could have been caused by this. Anyway, it seems
6590 much more sensible not to leave this file open at all, so that is what now
6593 29. The long-running daemon process does not normally write to the log once it
6594 has entered its main loop, and it closes the log before doing so. This is
6595 so that log files can straightforwardly be renamed and moved. However,
6596 there are a couple of unusual error situations where the daemon does write
6597 log entries, and I had neglected to close the log afterwards.
6599 30. The text of an SMTP error response that was received during a remote
6600 delivery was being truncated at 512 bytes. This is too short for some of
6601 the long messages that one sometimes sees. I've increased the limit to
6604 31. It is now possible to make retry rules that apply only when a message has a
6605 specific sender, in particular, an empty sender.
6607 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6608 it possible to be selective about when SMTP synchronization is enforced.
6610 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6612 32. Implemented hosts_connection_nolog.
6614 33. Added an ACL for QUIT.
6616 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6619 35. Added mailbox_size and mailbox_filecount to appendfile.
6621 36. Added control = no_multiline_responses to ACLs.
6623 37. There was a bug in the logic of the code that waits for the clock to tick
6624 in the case where the clock went backwards by a substantial amount such
6625 that the microsecond fraction of "now" was more than the microsecond
6626 fraction of "then" (but the whole seconds number was less).
6628 38. Added support for the libradius Radius client library this is found on
6629 FreeBSD (previously only the radiusclient library was supported).
6635 1. When certain lookups returned multiple values in the form name=value, the
6636 quoting of the values was not always being done properly. Specifically:
6637 (a) If the value started with a double quote, but contained no whitespace,
6639 (b) If the value contained whitespace other than a space character (i.e.
6640 tabs or newlines or carriage returns) it was not quoted.
6641 This fix has been applied to the mysql and pgsql lookups by writing a
6642 separate quoting function and calling it from the lookup code. The fix
6643 should probably also be applied to nisplus, ibase and oracle lookups, but
6644 since I cannot test any of those, I have not disturbed their existing code.
6646 2. A hit in the callout cache for a specific address caused a log line with no
6647 reason for rejecting RCPT. Now it says "Previous (cached) callout
6648 verification failure".
6650 3. There was an off-by-one bug in the queryprogram router. An over-long
6651 return line was truncated at 256 instead of 255 characters, thereby
6652 overflowing its buffer with the terminating zero. As well as fixing this, I
6653 have increased the buffer size to 1024 (and made a note to document this).
6655 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6656 when Exim is waiting for an SMTP response from a remote server, Exim
6657 restarts its select() call on the socket, thereby resetting its timeout.
6658 This is not a problem when such interrupts are rare. Somebody set up a cron
6659 job to run exiwhat every 2 minutes, which is less than the normal select()
6660 timeout (5 or 10 minutes). This meant that the select() timeout never
6661 kicked in because it was always reset. I have fixed this by comparing the
6662 time when an interrupt arrives with the time at the start of the first call
6663 to select(). If more time than the timeout has elapsed, the interrupt is
6664 treated as a timeout.
6666 5. Some internal re-factoring in preparation for the addition of Sieve
6667 extensions (by MH). In particular, the "personal" test is moved to a
6668 separate function, and given an option for scanning Cc: and Bcc: (which is
6669 not set for Exim filters).
6671 6. When Exim created an email address using the login of the caller as the
6672 local part (e.g. when creating a From: or Sender: header line), it was not
6673 quoting the local part when it contained special characters such as @.
6675 7. Installed new OpenBSD configuration files.
6677 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6678 try to make them clearer.
6680 9. Callout options, other than the timeout value, were being ignored when
6681 verifying sender addresses in header lines. For example, when using
6683 verify = header_sender/callout=no_cache
6685 the cache was (incorrectly) being used.
6687 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6688 only the Cygwin environment.
6690 11. When return_path_on_delivery was set as a log selector, if different remote
6691 addresses in the same message used different return paths and parallel
6692 remote delivery occurred, the wrong values would sometimes be logged.
6693 (Whenever a remote delivery process finished, the return path value from
6694 the most recently started remote delivery process was logged.)
6696 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6697 lines when AUTH and/or TLS are in use. This is the "received protocol"
6698 field. Exim used to use "asmtp" for authenticated SMTP, without any
6699 indication (in the protocol name) for TLS use. Now it follows the RFC and
6700 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6701 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6702 names appear in log lines as well as in Received: header lines.
6704 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6705 extensions, and comparison tests, and to fix some bugs.
6707 14. Changes to the "personal" filter test:
6709 (1) The test was buggy in that it was just doing the equivalent of
6710 "contains" tests on header lines. For example, if a user's address was
6711 anne@some.where, the "personal" test would incorrectly be true for
6713 To: susanne@some.where
6715 This test is now done by extracting each address from the header in turn,
6716 and checking the entire address. Other tests that are part of "personal"
6717 are now done using regular expressions (for example, to check local parts
6718 of addresses in From: header lines).
6720 (2) The list of non-personal local parts in From: addresses has been
6721 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6722 taken from the Sieve specification recommendations.
6724 (3) If the message contains any header line starting with "List-" it is
6725 treated as non-personal.
6727 (4) The test for "circular" in the Subject: header line has been removed
6728 because it now seems ill-conceived.
6730 15. Minor typos in src/EDITME comments corrected.
6732 16. Installed latest exipick from John Jetmore.
6734 17. If headers_add on a router specified a text string that was too long for
6735 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6736 of string_sprintf() is now avoided.
6738 18. $message_body_size was not set (it was always zero) when running the DATA
6739 ACL and the local_scan() function.
6741 19. For the "mail" command in an Exim filter, no default was being set for
6742 the once_repeat time, causing a random time value to be used if "once" was
6743 specified. (If the value happened to be <= 0, no repeat happened.) The
6744 default is now 0s, meaning "never repeat". The "vacation" command was OK
6745 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6746 (I found it when inspecting the code).
6748 20. There is now an overall timeout for performing a callout verification. It
6749 defaults to 4 times the callout timeout, which applies to individual SMTP
6750 commands during the callout. The overall timeout applies when there is more
6751 than one host that can be tried. The timeout is checked before trying the
6752 next host. This prevents very long delays if there are a large number of
6753 hosts and all are timing out (e.g. when the network connections are timing
6754 out). The value of the overall timeout can be changed by specifying an
6755 additional sub-option for "callout", called "maxwait". For example:
6757 verify = sender/callout=5s,maxwait=20s
6759 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6760 to the end before writing, but this should make it even safer).
6762 22. Exim was forgetting that it had advertised PIPELINING for the second and
6763 subsequent messages on an SMTP connection. It was also not resetting its
6764 memory on STARTTLS and an internal HELO.
6766 23. When Exim logs an SMTP synchronization error within a session, it now
6767 records whether PIPELINING has been advertised or not.
6769 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6770 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6771 rather than long int.
6773 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6775 26. Added the never_mail option to autoreply.
6781 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6782 crash if the getsockname() call failed; this can happen if a connection is
6783 closed very soon after it is established. The problem was simply in the
6784 order in which certain operations were done, causing Exim to try to write
6785 to the SMTP stream before it had set up the file descriptor. The bug has
6786 been fixed by making things happen in the correct order.
6792 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6793 before the connection was closed, thus losing the rejection response.
6795 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6796 some early Solaris releases, but causes trouble in current releases where
6797 socklen_t is defined.
6799 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6802 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6805 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6806 that releasing the top of it at the end releases what was used for sub-
6807 expansions (unless the block got too big). However, discard this block if
6808 the first thing is a variable or header, so that we can use its block when
6809 it is dynamic (useful for very large $message_headers, for example).
6811 6. Lookups now cache *every* query, not just the most recent. A new, separate
6812 store pool is used for this. It can be recovered when all lookup caches are
6813 flushed. Lookups now release memory at the end of their result strings.
6814 This has involved some general refactoring of the lookup sources.
6816 7. Some code has been added to the store_xxx() functions to reduce the amount
6817 of flapping under certain conditions.
6819 8. log_incoming_interface used to affect only the <= reception log lines. Now
6820 it causes the local interface and port to be added to several more SMTP log
6821 lines, for example "SMTP connection from", and rejection lines.
6823 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6825 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6827 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6828 could be overwritten at the end of the current message (or the start of a
6829 new message if it was set in a HELO ACL). The value is now preserved for
6830 the duration of the SMTP connection.
6832 12. If a transport had a headers_rewrite setting, and a matching header line
6833 contained an unqualified address, that address was qualified, even if it
6834 did not match any rewriting rules. The underlying bug was that the values
6835 of the flags that permit the existence of unqualified sender and recipient
6836 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6837 non-local messages, and by -bnq for local messages) were not being
6838 preserved with the message after it was received.
6840 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6841 "next input=" as part of the text comprising the host identity instead of
6842 the correct text. The code was using the same buffer for two different
6843 strings. However, depending on which order the printing function evaluated
6844 its arguments, the bug did not always show up. Under Linux, for example, my
6845 test suite worked just fine.
6847 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6848 doesn't work with some older versions of Perl. It has been changed to "my",
6849 which in any case is probably the better facility to use.
6851 15. A really picky compiler found some instances of statements for creating
6852 error messages that either had too many or two few arguments for the format
6855 16. The size of the buffer for calls to the DNS resolver has been increased
6856 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6857 for addresses that have a lot of PTR records. This alleviates a problem; it
6858 does not fully solve it.
6860 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6861 buffer now truncates the list and logs the incident, which is the same
6862 action as happens when Exim is looking up a host name and its aliases.
6863 Previously in this situation something unpredictable would happen;
6864 sometimes it was "internal error: store_reset failed".
6866 18. If a server dropped the connection unexpectedly when an Exim client was
6867 using GnuTLS and trying to read a response, the client delivery process
6868 crashed while trying to generate an error log message.
6870 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6871 string, for example:
6873 warn message = H1: something\nH2: something
6875 the text was added as a single header line from Exim's point of view
6876 though it ended up OK in the delivered message. However, searching for the
6877 second and subsequent header lines using $h_h2: did not work. This has been
6878 fixed. Similarly, if a system filter added multiple headers in this way,
6879 the routers could not see them.
6881 20. Expanded the error message when iplsearch is called with an invalid key to
6882 suggest using net-iplsearch in a host list.
6884 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6885 ACLs are no longer actually imposed (and a message to that effect is
6888 22. If a "gecos" field in a passwd entry contained escaped characters, in
6889 particular, if it contained a \" sequence, Exim got it wrong when building
6890 a From: or a Sender: header from that name. A second bug also caused
6891 incorrect handling when an unquoted " was present following a character
6892 that needed quoting.
6894 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6895 was not being matched caselessly.
6897 24. Arranged for all hyphens in the exim.8 source to be escaped with
6900 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6901 properly. Recipient callout cache records were still being keyed to include
6902 the sender, even when use_sender was set false. This led to far more
6903 callouts that were necessary. The sender is no longer included in the key
6904 when use_sender is false.
6906 26. Added "control = submission" modifier to ACLs.
6908 27. Added the ${base62d: operator to decode base 62 numbers.
6910 28. dnsdb lookups can now access SRV records.
6912 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6913 the configuration file.
6915 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6916 (-v) mode. This makes the output for a verbose queue run more intelligible.
6918 31. Added a use_postmaster feature to recipient callouts.
6920 32. Added the $body_zerocount variable, containing the number of binary zero
6921 bytes in the message body.
6923 33. The time of last modification of the "new" subdirectory is now used as the
6924 "mailbox time last read" when there is a quota error for a maildir
6927 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6929 35. Added +ignore_unknown as a special item in host lists.
6931 36. Code for decoding IPv6 addresses in host lists is now included, even if
6932 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6933 address was recognized as an IP address, but was then not correctly decoded
6934 into binary, causing unexpected and incorrect effects when compared with
6941 1. Very minor rewording of debugging text in manualroute to say "list of
6942 hosts" instead of "hostlist".
6944 2. If verify=header_syntax was set, and a header line with an unqualified
6945 address (no domain) and a large number of spaces between the end of the
6946 name and the colon was received, the reception process suffered a buffer
6947 overflow, and (when I tested it) crashed. This was caused by some obsolete
6948 code that should have been removed. The fix is to remove it!
6950 3. When running in the test harness, delay a bit after writing a bounce
6951 message to get a bit more predictability in the log output.
6953 4. Added a call to search_tidyup() just before forking a reception process. In
6954 theory, someone could use a lookup in the expansion of smtp_accept_max_
6955 per_host which, without the tidyup, could leave open a database connection.
6957 5. Added the variables $recipient_data and $sender_data which get set from a
6958 lookup success in an ACL "recipients" or "senders" condition, or a router
6959 "senders" option, similar to $domain_data and $local_part_data.
6961 6. Moved the writing of debug_print from before to after the "senders" test
6964 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6965 problems for message scanning, either using a data ACL, or using
6966 local_scan() because the Received: header was not generated till after they
6967 were called (in order to set the time as the time of reception completion).
6968 I have revised the way this works. The header is now generated after the
6969 body is received, but before the ACL or local_scan() are called. After they
6970 are run, the timestamp in the header is updated.
6976 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6977 before starting a queue runner without re-exec. This happened only when
6978 deliver_drop_privilege was set or when the Exim user was set to root. The
6979 effect of the bug was that timeouts during subsequent deliveries caused
6980 crashes instead of being properly handled. The handler is now left at its
6981 default (and expected) setting.
6983 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6984 message, again when deliver_drop_privilege is set or Exim is run as root.
6985 The bug described in (1) was not present in this case, but the tidying up
6986 of the other signals was missing. I have made the two cases consistent.
6988 3. The ignore_target_hosts setting on a manualroute router was being ignored
6989 for hosts that were looked up using the /MX notation.
6991 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6994 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6995 operated on the sender address. After changing the $sender_address to <>
6996 for the sender address verify, Exim was re-instated it as the original
6997 (before rewriting) address, but remembering that it had rewritten it, so it
6998 wasn't rewriting it again. This bug also had the effect of breaking the
6999 sender address verification caching when the sender address was rewritten.
7001 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7002 This has been changed so that if the ip literal address matches
7003 ignore_target_hosts, the router declines.
7005 7. Added expansion conditions match_domain, match_address, and match_local_
7006 part (NOT match_host).
7008 8. The placeholder for the Received: header didn't have a length field set.
7010 9. Added code to Exim itself and to exim_lock to test for a specific race
7011 condition that could lead to file corruption when using MBX delivery. The
7012 issue is with the lockfile that is created in /tmp. If this file is removed
7013 after a process has opened it but before that process has acquired a lock,
7014 there is the potential for a second process to recreate the file and also
7015 acquire a lock. This could lead to two Exim processes writing to the file
7016 at the same time. The added code performs the same test as UW imapd; it
7017 checks after acquiring the lock that its file descriptor still refers to
7018 the same named file.
7020 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7021 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7024 11. Added the smtp_active_hostname option. If used, this will typically be made
7025 to depend on the incoming interface address. Because $interface_address is
7026 not set up until the daemon has forked a reception process, error responses
7027 that can happen earlier (such as "too many connections") no longer contain
7030 12. If an expansion in a condition on a "warn" statement fails because a lookup
7031 defers, the "warn" statement is abandoned, and the next ACL statement is
7032 processed. Previously this caused the whole ACL to be aborted.
7034 13. Added the iplsearch lookup type.
7036 14. Added ident_timeout as a log selector.
7038 15. Added tls_certificate_verified as a log selector.
7040 16. Added a global option tls_require_ciphers (compare the smtp transport
7041 option of the same name). This controls incoming TLS connections.
7043 17. I finally figured out how to make tls_require_ciphers do a similar thing
7044 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7045 before starting the TLS session.
7047 18. Tabs are now shown as \t in -bP output.
7049 19. If the log selector return_path_on_delivery was set, Exim crashed when
7050 bouncing a message because it had too many Received: header lines.
7052 20. If two routers both had headers_remove settings, and the first one included
7053 a superfluous trailing colon, the final name in the first list and the
7054 first name in the second list were incorrectly joined into one item (with a
7055 colon in the middle).
7061 1. Added -C and -D options to the exinext utility, mainly to make it easier
7062 to include in the automated testing, but these could be helpful when
7063 multiple configurations are in use.
7065 2. The exinext utility was not formatting the output nicely when there was
7066 an alternate port involved in the retry record key, nor when there was a
7067 message id as well (for retries that were specific to a specific message
7068 and a specific host). It was also confused by IPv6 addresses, because of
7069 the additional colons they contain. I have fixed the IPv4 problem, and
7070 patched it up to do a reasonable job for IPv6.
7072 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7073 delivery, the log line now contains "pipelined" if PIPELINING was used.
7075 4. An SMTP transport process used to panic and die if the bind() call to set
7076 an explicit outgoing interface failed. This has been changed; it is now
7077 treated in the same way as a connect() failure.
7079 5. A reference to $sender_host_name in the part of a conditional expansion
7080 that was being skipped was still causing a DNS lookup. This no longer
7083 6. The def: expansion condition was not recognizing references to header lines
7084 that used bh_ and bheader_.
7086 7. Added the _cache feature to named lists.
7088 8. The code for checking quota_filecount in the appendfile transport was
7089 allowing one more file than it should have been.
7091 9. For compatibility with Sendmail, the command line option
7099 and sets the incoming protocol and host name (for trusted callers). The
7100 host name and its colon can be omitted when only the protocol is to be set.
7101 Note the Exim already has two private options, -pd and -ps, that refer to
7102 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7103 "s", but I don't think that's a major issue.
7105 10. A number of refactoring changes to the code, none of which should affect
7108 (a) The number of logging options was getting close to filling up the
7109 32-bit word that was used as a bit map. I have split them into two classes:
7110 those that are passed in the argument to log_write(), and those that are
7111 only ever tested independently outside of that function. These are now in
7112 separate 32-bit words, so there is plenty of room for expansion again.
7113 There is no change in the user interface or the logging behaviour.
7115 (b) When building, for example, log lines, the code previously used a
7116 macro that called string_cat() twice, in order to add two strings. This is
7117 not really sufficiently general. Furthermore, there was one instance where
7118 it was actually wrong because one of the argument was used twice, and in
7119 one call a function was used. (As it happened, calling the function twice
7120 did not affect the overall behaviour.) The macro has been replaced by a
7121 function that can join an arbitrary number of extra strings onto a growing
7124 (c) The code for expansion conditions now uses a table and a binary chop
7125 instead of a serial search (which was left over from when there were very
7126 few conditions). Also, it now recognizes conditions like "pam" even when
7127 the relevant support is not compiled in: a suitably worded error message is
7128 given if an attempt is made to use such a condition.
7130 11. Added ${time_interval:xxxxx}.
7132 12. A bug was causing one of the ddress fields not to be passed back correctly
7133 from remote delivery subprocesses. The field in question was not being
7134 subsequently used, so this caused to problems in practice.
7136 13. Added new log selectors queue_time and deliver_time.
7138 14. Might have fixed a bug in maildirsizefile handling that threw up
7139 "unexpected character" debug warnings, and recalculated the data
7140 unnecessarily. In any case, I expanded the warning message to give more
7143 15. Added the message "Restricted characters in address" to the statements in
7144 the default ACL that block characters like @ and % in local parts.
7146 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7147 Three changes have been made:
7149 (a) There was a serious bug; a negative response to MAIL caused the whole
7150 recipient domain to be cached as invalid, thereby blocking all messages
7151 to all local parts at the same domain, from all senders. This bug has
7152 been fixed. The domain is no longer cached after a negative response to
7153 MAIL if the sender used is not empty.
7155 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7158 (c) A new callout option, "use_sender" has been added for people who want
7159 the modified behaviour.
7165 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7168 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7169 indeed breaks things for older releases.
7171 3. Added additional logging to the case where there is a problem reading data
7172 from a filter that is running in a subprocess using a pipe, in order to
7173 try to track down a specific problem.
7175 4. Testing facility fudge: when running in the test harness and attempting
7176 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7177 getting "No route to host". Convert this to a timeout.
7179 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7182 6. Some OS don't have socklen_t but use size_t instead. This affects the
7183 fifth argument of getsockopt() amongst other things. This is now
7184 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7185 can be set for individual OS. I have set it for SunOS5, OSF1, and
7186 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7187 some earlier ones do not.
7189 7. Change 4.30/15 was not doing the test caselessly.
7191 8. The standard form for an IPv6 address literal was being rejected by address
7192 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7193 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7194 this, as well as the form without the "IPv6" on the front (but only when
7195 address literals are enabled, of course).
7197 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7199 10. Exim crashed if a message with an empty sender address specified by -f
7200 encountered a router with an errors_to setting. This could be provoked only
7201 by a command such as
7205 where an empty string was supplied; "<>" did not hit this bug.
7207 11. Installed PCRE release 4.5.
7209 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7210 remained set. It is now erased.
7212 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7213 times from message ids (which are base 36 rather than the normal 62).
7215 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7216 were being counted as actual protocol errors, and logged if the log
7217 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7218 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7219 and DATA following a set of rejected RCPTs do not count as protocol errors.
7220 In other words, Exim assumes they were pipelined, though this may not
7221 actually be the case. Of course, in all cases the client gets an
7222 appropriate error code.
7224 15. If a lookup fails in an ACL condition, a message about the failure may
7225 be available; it is used if testing the ACL cannot continue, because most
7226 such messages specify what the cause of the deferral is. However, some
7227 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7228 that caused an old message to be retained and used if a later statement
7229 caused a defer, replacing the real cause of the deferral.
7231 16. If an IP address had so many PTR records that the DNS lookup buffer
7232 was not large enough to hold them, Exim could crash while trying to process
7233 the truncated data. It now detects and logs this case.
7235 17. Further to 4.21/58, another change has been made: if (and only if) the
7236 first line of a message (the first header line) ends with CRLF, a bare LF
7237 in a subsequent header line has a space inserted after it, so as not to
7238 terminate the header.
7240 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7241 unnecessarily, used atoi() instead of strtol(), and didn't check the
7242 termination when getting file sizes from file names by regex.
7244 19. Completely re-implemented the support for maildirsize files, in the light
7245 of a number of problems with the previous contributed implementation
7246 (4.30/29). In particular:
7248 . If the quota is zero, the maildirsize file is maintained, but no quota is
7251 . If the maildir directory does not exist, it is created before any attempt
7252 to write a maildirsize file.
7254 . The quota value in the file is just a cache; if the quota is changed in
7255 the transport, the new value overrides.
7257 . A regular expression is available for excluding directories from the
7260 20. The autoreply transport checks the characters in options that define the
7261 message's headers; it allows continued headers, but it was checking with
7262 isspace() after an embedded newline instead of explicitly looking for a
7265 21. If all the "regular" hosts to which an address was routed had passed their
7266 expiry times, and had not reached their retry times, the address was
7267 bounced, even if fallback hosts were defined. Now Exim should go on to try
7270 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7271 equivalent code in the SMTP transport. Some hosts send humungous responses
7272 to HELO/EHLO, more than 1024 it seems.
7274 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7275 gives compiler warnings in some environments. I've now done it "properly",
7278 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7279 (because of problems with the built-in one) was declared to return uschar *
7280 instead of char *, causing compiler failure.
7282 25. Fixed a file descriptor leak when processing alias/forward files.
7284 26. Fixed a minor format string issue in dbfn.c.
7286 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7288 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7289 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7292 29. When a sender address is verified, it is cached, to save repeating the test
7293 when there is more than one recipient in a message. However, when the
7294 verification involves a callout, it is possible for different callout
7295 options to be set for different recipients. It is too complicated to keep
7296 track of this in the cache, so now Exim always runs a verification when a
7297 callout is required, relying on the callout cache for the optimization.
7298 The overhead is duplication of the address routing, but this should not be
7301 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7302 to be verified with callout=postmaster, and the main callout worked but the
7303 postmaster check failed, the verification correctly failed. However, if a
7304 subsequent RCPT command asked for sender verification *without* the
7305 postmaster check, incorrect caching caused this verification also to fail,
7308 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7309 it was not caching the DNS options (qualify_single, search_parents) that
7310 were used when the lookup failed. A subsequent lookup with different
7311 options therefore always gave the same answer, though there were cases
7312 where it should not have. (Example: a "domains = !$mx_any" option on a
7313 dnslookup router: the "domains" option is always processed without any
7314 widening, but the router might have qualify_single set.) Now Exim uses the
7315 cached value only when the same options are set.
7317 32. Added John Jetmore's "exipick" utility to the distribution.
7319 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7320 than a timeout (e.g. a certificate is required, and is not provided), an
7321 Exim server now closes the connection immediately. Previously it waited for
7322 the client to close - but if the client is SSL, it seems that they each
7323 wait for each other, leading to a delay before one of them times out.
7325 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7326 maintained 0.8.x compatibility because I don't think many are using it, and
7327 it is clearly obsolete.
7329 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7332 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7333 client certificate was expired. A simple patch fixes this, though I don't
7334 understand the full logic of why the verify callback is called multiple
7337 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7338 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7339 which causes problems with some clients (such as the Certicom SSL Plus
7340 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7341 disables the coutermeasure allowing Eudora to connect."
7343 38. Exim was not checking that a write() to a log file succeeded. This could
7344 lead to Bad Things if a log got too big, in particular if it hit a file
7345 size limit. Exim now panics and dies if it cannot write to a log file, just
7346 as it does if it cannot open a log file.
7348 39. Modified OS/Makefile-Linux so that it now contains
7350 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7352 The two -D definitions ensure that Exim is compiled with large file
7353 support, which makes it possible to handle log files that are bigger than
7356 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7357 instance) a domain was checked against a named list that involved a lookup,
7358 causing $domain_data to be set, then another domain was checked against the
7359 same list, then the first domain was re-checked, the value of $domain_data
7360 after the final check could be wrong. In particular, if the second check
7361 failed, it could be set empty. This bug probably also applied to
7364 41. The strip_trailing_dot option was not being applied to the address given
7365 with the -f command-line option.
7367 42. The code for reading a message's header from the spool was incrementing
7368 $received_count, but never initializing it. This meant that the value was
7369 incorrect (doubled) while delivering a message in the same process in which
7370 it was received. In the most common configuration of Exim, this never
7371 happens - a fresh exec is done - but it can happen when
7372 deliver_drop_privilege is set.
7374 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7375 it now includes up to 150 characters of the unexpected data in the log
7378 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7379 and building data strings. The size of both of these buffers was 10 000
7380 bytes - far larger than anybody would *ever* want, thought I. Needless to
7381 say, somebody hit the limit. I have increased the maximum line length to
7382 20 000 and the maximum data length of concatenated lines to 100 000. I have
7383 also fixed two bugs, because there was no checking on these buffers. Tsk,
7384 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7385 buffer is too small.
7387 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7388 lsearch lookups. Now it does.
7390 46. When parsing a route_list item in a manualroute router, a fixed-length
7391 buffer was used for the list of hosts. I made this 1024 bytes long,
7392 thinking that nobody would ever have a list of hosts that long. Wrong.
7393 Somebody had a whole pile of complicated expansion conditions, and the
7394 string was silently truncated, leading to an expansion error. It turns out
7395 that it is easier to change to an unlimited length (owing to other changes
7396 that have happened since this code was originally written) than to build
7397 structure for giving a limitation error. The length of the item that
7398 expands into the list of hosts is now unlimited.
7400 47. The lsearch lookup could not handle data where the length of text line was
7401 more than 4095 characters. Such lines were truncated, leading to shortened
7402 data being returned. It should now handle lines of any length.
7404 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7405 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7408 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7409 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7410 no longer happens in comment lines. A list of replacements is now placed
7411 at the head of all of the source files, except those whose only change is
7412 to replace PERL_COMMAND in the very first #! line.
7414 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7415 messages on the queue, with a bottom-up merge sort, using code contributed
7416 by Michael Haardt. This should make operations like -bp somewhat faster on
7417 large queues. It won't affect queue runners, except when queue_run_in_order
7420 51. Installed eximstats 1.31 in the distribution.
7422 52. Added support for SRV lookups to the dnslookup router.
7424 53. If an ACL referred to $message_body or $message_body_end, the value was not
7425 reset for any messages that followed in the same SMTP session.
7427 54. The store-handling optimization for building very long strings was not
7428 differentiating between the different store pools. I don't think this
7429 actually made any difference in practice, but I've tidied it.
7431 55. While running the routers to verify a sender address, $sender_address
7432 was still set to the sender address. This is wrong, because when routing to
7433 send a bounce to the sender, it would be empty. Therefore, I have changed
7434 it so that, while verifying a sender address, $sender_address is set to <>.
7435 (There is no change to what happens when verifying a recipient address.)
7437 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7438 target A or AAAA records (if not already returned) without resetting the
7439 qualify_single or search_parents options of the DNS resolver. These are
7440 inappropriate in this case because the targets of MX and SRV records must
7441 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7442 target that, when qualified, matched something in the local domain. These
7443 two options are now turned off when doing these lookups.
7445 57. It seems that at least some releases of Reiserfs (which does not have the
7446 concept of a fixed number of inodes) returns zero and not -1 for the
7447 number of available inodes. This interacted badly with check_spool_inodes,
7448 which assumed that -1 was the "no such thing" setting. What I have done is
7449 to check that the total number of inodes is greater than zero before doing
7450 the test of how many are available.
7452 58. When a "warn" ACL statement has a log_message modifier, the message is
7453 remembered, and not repeated. This is to avoid a lot of repetition when a
7454 message has many recipients that cause the same warning to be written.
7455 However, Exim was preserving the list of already written lines for an
7456 entire SMTP session, which doesn't seem right. The memory is now reset if a
7457 new message is started.
7459 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7460 debugging output unless log_rewrite was also set.
7462 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7463 of (int)(handle) when we know that handle contains (void *)(-1).
7465 61. The Exim daemon panic-logs an error return when it closes the incoming
7466 connection. However "connection reset by peer" seems to be common, and
7467 isn't really an error worthy of noting specially, so that particular error
7470 62. When Exim is trying to find all the local interfaces, it used to panic and
7471 die if the ioctl to get the interface flags failed. However, it seems that
7472 on at least one OS (Solaris 9) it is possible to have an interface that is
7473 included in the list of interfaces, but for which you get a failure error
7474 for this call. This happens when the interface is not "plumbed" into a
7475 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7476 failure of the "get flags" call assumes that the interface is down.
7478 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7479 makes life easier for people who are doing arithmetic on fields extracted
7480 from dates, where you often get leading zeros that should not be
7481 interpreted as octal.
7483 64. Added qualify_domain to the redirect router, to override the global
7486 65. If a pathologically long header line contained very many addresses (the
7487 report of this problem mentioned 10 000) and each of them was rewritten,
7488 Exim could use up a very large amount of memory. (It kept on making new
7489 copies of the header line as it rewrote, and never released the old ones.)
7490 At the expense of a bit more processing, the header rewriting function has
7491 been changed so that it no longer eats memory in this way.
7493 66. The generation of the Received: header has been moved from the time that a
7494 message starts to be received, to the time that it finishes. The timestamp
7495 in the Received: header should now be very close to that of the <= log
7496 line. There are two side-effects of this change:
7498 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7499 logged header lines no longer include the local Received: line, because
7500 it has not yet been created. The same applies to a copy of the message
7501 that is returned to a non-SMTP sender when a message is rejected.
7503 (b) When a filter file is tested using -bf, no additional Received: header
7504 is added to the test message. After some thought, I decided that this
7507 This change does not affect the value of $received_for. It is still set
7508 after address rewriting, but before local_scan() is called.
7510 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7512 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7513 gave an unhelpful panic error message, and a defer error. I have managed to
7514 change this behaviour so that it now rejects any supplied certificate,
7515 which seems right, as the list of acceptable certificates is empty.
7517 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7518 gave an unhelpful defer error. I have not managed to make this reject any
7519 supplied certificates, but the error message it gives is "no certificate
7520 supplied", which is not helpful.
7522 70. exigrep's output now also includes lines that are not associated with any
7523 message, but which match the given pattern. Implemented by a patch from
7524 Martin Sluka, which also tidied up the Perl a bit.
7526 71. Recipient callout verification, like sender verification, was using <> in
7527 the MAIL FROM command. This isn't really the right thing, since the actual
7528 sender may affect whether the remote host accepts the recipient or not. I
7529 have changed it to use the actual sender in the callout; this means that
7530 the cache record is now keyed on a recipient/sender pair, not just the
7531 recipient address. There doesn't seem to be a real danger of callout loops,
7532 since a callout by the remote host to check the sender would use <>.
7533 [SEE ABOVE: changed after hitting problems.]
7535 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7536 temporary errors. However, in the case of such a code being given after
7537 the end of a data transmission (i.e. after ".") Exim was failing to write
7538 a retry record for the message. (Yes, there was some broken host that was
7539 actually sending 8xx at this point.)
7541 73. An unknown lookup type in a host list could cause Exim to panic-die when
7542 the list was checked. (An example that provoked this was putting <; in the
7543 middle of a list instead of at the start.) If this happened during a DATA
7544 ACL check, a -D file could be left lying around. This kind of configuration
7545 error no longer causes Exim to die; instead it causes a defer error. The
7546 incident is still logged to the main and panic logs.
7548 74. Buglet left over from Exim 3 conversion. The message "too many messages
7549 in one connection" was written to the rejectlog but not the mainlog, except
7550 when address rewriting (yes!) was being logged.
7552 75. Added write_rejectlog option.
7554 76. When a system filter was run not as root (that is, when system_filter_user
7555 was set), the values of the $n variables were not being returned to the
7556 main process; thus, they were not subsequently available in the $sn
7559 77. Added +return_path_on_delivery log selector.
7561 78. A connection timeout was being treated differently from recipients deferred
7562 when testing hosts_max_try with a message that was older than the host's
7563 retry timeout. (The host should not be counted, thus allowing all hosts to
7564 be tried at least once before bouncing.) This may have been the cause of an
7565 occasionally reported bug whereby a message would remain on the queue
7566 longer than the retry timeout, but would be bounced if a delivery was
7567 forced. I say "may" because I never totally pinned down the problem;
7568 setting up timeout/retry tests is difficult. See also the next item.
7570 79. The ultimate address timeout was not being applied to errors that involved
7571 a combination of host plus message (for example, a timeout on a MAIL
7572 command). When an address resolved to a number of possible hosts, and they
7573 were not all tried for each delivery (e.g. because of hosts_max_try), a
7574 message could remain on the queue longer than the retry timeout.
7576 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7579 81. Fixed an obscure SMTP outgoing bug which required at least the following
7580 conditions: (a) there was another message waiting for the same server;
7581 (b) the server returned 5xx to all RCPT commands in the first message so
7582 that the message was not completed; (c) the server dropped the connection
7583 or gave a negative response to the RSET that Exim sends to abort the
7584 transaction. The observed case was a dropped connection after DATA that had
7585 been sent in pipelining mode. That is, the server had advertised PIPELINING
7586 but was not implementing it correctly. The effect of the bug was incorrect
7587 behaviour, such as trying another host, and this could lead to a crash.
7593 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7594 and daemon.c were passed as pointers to ints; they should have been
7595 pointers to socklen_t variables (which are typically unsigned ints).
7597 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7600 3. Fixed a really odd bug that affected only the testing scheme; patching a
7601 certain fixed string in the binary changed the value of another string that
7602 happened to be identical to the end of the original first string.
7604 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7605 name", it returns that address as the IP address. On some operating
7606 systems (e.g. Solaris), it also passes back the IP address string as the
7607 "host name". However, on others (e.g. Linux), it passes back an empty
7608 string. Exim wasn't checking for this, and was changing the host name to an
7609 empty string, assuming it had been canonicalized.
7611 5. Although rare, it is permitted to have more than one PTR record for a given
7612 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7613 all the names associated with an address, because they do in Solaris.
7614 However, it seems that they do not in Linux for data that comes from the
7615 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7616 I found this out when I moved to a new Linux workstation and tried to run
7617 the Exim test suite.
7619 To get round this problem I have changed the code so that it now does its
7620 own call to the DNS to look up PTR records when searching for a host name.
7621 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7622 addresses that are only in /etc/hosts are still found.
7624 This behaviour is, however, controlled by an option called host_lookup_
7625 order, which defaults to "bydns:byaddr". If people want to use the other
7626 order, or indeed, just use one or the other means of lookup, they can
7627 specify it in this variable.
7629 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7630 some operating systems, this comes back from gethostbyaddr() as an empty
7631 string, and this is what Exim used to test for. However, it seems that in
7632 other systems, "." is yielded. Exim now tests for this case too.
7634 7. The values of check_spool_space and check_log_space are now held internally
7635 as a number of kilobytes instead of an absolute number of bytes. If a
7636 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7637 kilobyte. This means that much larger values can be stored.
7639 8. Exim monitor: an attempt to get the action menu when not actually pointing
7640 at a message produces an empty menu entitled "No message selected". This
7641 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7642 no entries in it ("Shell widget menu has zero width and/or height"). So I
7643 have added a single, blank menu entry in this case.
7645 9. Added ${quote_local_part.
7647 10. MIME decoding is now applied to the contents of Subject: header lines when
7650 11. Now that a reference to $sender_host_address automatically causes a reverse
7651 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7652 host lookup before query-style lookups in lists that might use this
7653 variable. This has therefore been abolished, and the "net-" prefix is no
7654 longer necessary for query-style lookups.
7656 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7657 have been a typo for LFLAGS, so it has been changed.
7659 13. The install script calls Exim with "-C /dev/null" in order to find the
7660 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7661 to be output. However, since Exim outputs its version number before the
7662 error, it didn't break the script. It just looked ugly. I fixed this by
7663 always allowing "-C /dev/null" if the caller is root.
7665 14. Ignore overlarge ACL variable number when reading spool file - insurance
7666 against a later release with more variables having written the file.
7668 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7669 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7670 this, as well as the form without the "IPv6" on the front.
7672 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7673 OS/Makefile-Darwin file.
7675 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7676 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7678 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7679 variable. However, some OS use ints for inodes. I've added cast to long int
7680 to get rid of the compiler warning.
7682 19. I had forgotten to lock out "/../" in configuration file names when
7683 ALT_CONFIG_PREFIX was set.
7685 20. Routers used for verification do not need to specify transports. However,
7686 if such a router generated a host list, and callout was configured, Exim
7687 crashed, because it could not find a port number from the (non-existent)
7688 transport. It now assumes port 25 in this circumstance.
7690 21. Added the -t option to exigrep.
7692 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7693 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7694 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7695 them is set, LOOKUP_LSEARCH is forced.
7697 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7699 24. Added sender and host information to the "rejected by local_scan()" log
7700 line; previously there was no indication of these.
7702 25. Added .include_if_exists.
7704 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7705 Linux. It turns out that not all file systems support this. Apparently some
7706 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7707 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7708 not supported on the file descriptor, is now ignored when Exim is trying to
7709 sync a directory. This applies only to Linux.
7711 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7713 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7714 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7717 29. Added support for maildirsize files from supplied patch (modified a bit).
7719 30. The use of :fail: followed by an empty string could lead Exim to respond to
7720 sender verification failures with (e.g.):
7722 550 Verification failed for <xxx>
7723 550 Sender verify failed
7725 where the first response line was missing the '-' that indicates it is not
7726 the final line of the response.
7728 31. The loop for finding the name of the user that called Exim had a hardwired
7729 limit of 10; it now uses the value of finduser_retries, which is used for
7730 all other user lookups.
7732 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7735 33. Exim was neglecting to zero errno before one call of strtol() when
7736 expanding a string and expecting an integer value. On some systems this
7737 resulted in spurious "integer overflow" errors. Also, it was casting the
7738 result into an int without checking.
7740 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7741 did not work. The code looks as if it has *never* worked, though it appears
7742 to have been documented since at least release 1.62. I have made it work.
7744 35. The "timeout_DNS" error in retry rules, also documented since at least
7745 1.62, also never worked. As it isn't clear exactly what this means, and
7746 clearly it isn't a major issue, I have abolished the feature by treating it
7747 as "timeout", and writing a warning to the main and panic logs.
7749 36. The display of retry rules for -brt wasn't always showing the error code
7752 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7753 timeout_connect_A, timeout_connect_MX.
7755 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7756 to the empty sender.
7758 39. The daemon was not analysing the content of -oX till after it had closed
7759 stderr and disconnected from the controlling terminal. This meant that any
7760 syntax errors were only noted on the panic log, and the return code from
7761 the command was 0. By re-arranging the code a little, I've made the
7762 decoding happen first, so such errors now appear on stderr, and the return
7763 code is 1. However, the actual setting up of the sockets still happens in
7764 the disconnected process, so errors there are still only recorded on the
7767 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7768 connections (as happens on some IP stacks) was logged at start up time as
7769 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7770 it from "IPv6 and IPv4", which means that two separate sockets are being
7773 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7774 says whether AF_INET or AF_INET6 was passed as an argument.
7776 42. Exiwhat output was messed up when time zones were included in log
7779 43. Exiwhat now gives more information about the daemon's listening ports,
7780 and whether -tls-on-connect was used.
7782 44. The "port" option of the smtp transport is now expanded.
7784 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7785 silently ignored. Now an error message is written to the main and panic
7788 46. There's a new ACL modifier called "logwrite" which writes to a log file
7789 as soon as it is encountered.
7791 47. Added $local_user_uid and $local_user_gid at routing time.
7793 48. Exim crashed when trying to verify a sender address that was being
7796 49. Exim was recognizing only a space character after ".include". It now also
7797 recognizes a tab character.
7799 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7800 extracting the relevant information from the specification. The man page no
7801 longer contains scrambled data for the -d option, and I've added a section
7802 at the front about calling Exim under different names.
7804 51. Added "extra_headers" argument to the "mail" command in filter files.
7806 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7809 53. Installed eximstats 1.29.
7811 54. Added transport_filter_timeout as a generic transport option.
7813 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7814 Cc: header lines. This was required by RFC 822, but it not required by RFC
7817 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7818 incoming messages that did not have them. Now it does so only if the
7819 message originates locally, that is, if there is no associated remote host
7820 address. When Resent- header lines are present, this applies to the Resent-
7821 lines rather than the non-Resent- lines.
7823 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7824 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7826 58. Messages for configuration errors now include the name of the main
7827 configuration file - useful now that there may be more than one file in a
7828 list (.included file names were always shown).
7830 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7831 for those rare installations that do not start the daemon as root or run it
7832 setuid root. I've cut out the call to initgroups() if the daemon is not
7835 60. The Exim user and group can now be bound into the binary as text strings
7836 that are looked up at the start of Exim's processing.
7838 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7840 62. Added $mailstore_basename variable.
7842 63. Installed patch to sieve.c from Michael Haardt.
7844 64. When Exim failed to open the panic log after failing to open the main log,
7845 the original message it was trying to log was written to stderr and debug
7846 output, but if they were not available (the usual case in production), it
7847 was lost. Now it is written to syslog before the two lines that record the
7848 failures to open the logs.
7850 65. Users' Exim filters run in subprocesses under the user's uid. It is
7851 possible for a "deliver" command or an alias in a "personal" command to
7852 provoke an address rewrite. If logging of address rewriting is configured,
7853 this fails because the process is not running as root or exim. There may be
7854 a better way of dealing with this, but for the moment (because 4.30 needs
7855 to be released), I have disabled address rewrite logging when running a
7856 filter in a non-root, non-exim process.
7862 1. The buildconfig auxiliary program wasn't quoting the value set for
7863 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7864 not defined. This bug was masked in 4.22 by the effect that was fixed in
7867 2. Some messages that were rejected after a message id was allocated were
7868 shown as "incomplete" by exigrep. It no longer does this for messages that
7869 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7871 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7872 have allow_domain_literals set, the ID did not get logged in the <= line.
7873 Domain literals are now always recognized in Message-ID: header lines.
7875 4. The first argument for a ${extract expansion item is the key name or field
7876 number. Leading and trailing spaces in this item were not being ignored,
7877 causing some misleading effects.
7879 5. When deliver_drop_privilege was set, single queue runner processes started
7880 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7881 same command in the process it spins off) were not dropping privilege.
7883 6. When the daemon running as "exim" started a queue runner, it always
7884 re-executed Exim in the spun-off process. This is a waste of effort when
7885 deliver_drop_privilege is set. The new process now just calls the
7886 queue-runner function directly.
7892 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7895 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7896 ignored. Though the use of -oP was forcing the writing of a pid file, it
7897 was always written to the default place.
7899 3. If the message "no IP address found for host xxxx" is generated during
7900 incoming verification, it is now followed by identification of the incoming
7901 connection (so you can more easily find what provoked it).
7903 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7905 5. Added some features to "harden" Exim a bit more against certain attacks:
7907 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7908 be put in Local/Makefile. This is like the never_users runtime option,
7909 but it cannot be overridden. The default setting is "root".
7911 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7912 prefix string with which any file named in a -C command line option
7915 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7916 is retained for -C and -D only if the caller of Exim is root. Without
7917 it, the exim user may also use -C and -D and retain privilege.
7919 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7920 command line option is disabled.
7922 6. Macro names set by the -D option must start with an upper case letter, just
7923 like macro names defined in the configuration file.
7925 7. Added "dereference=" facility to LDAP.
7927 8. Two instances of the typo "uknown" in the source files are fixed.
7929 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7930 the Configure-Makefile script screwed up while processing it.
7932 10. Incorporated PCRE 4.4.
7934 11. The SMTP synchronization check was not operating right at the start of an
7935 SMTP session. For example, it could not catch a HELO sent before the client
7936 waited for the greeting. There is now a check for outstanding input at the
7937 point when the greeting is written. Because of the duplex, asynchronous
7938 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7939 way, but not yet received, when the check is performed.
7941 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7942 on TCP/IP sockets, because this apparently causes some broken clients to
7945 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7946 unchanged) from the Cygwin maintainer.
7948 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7949 was supported instead of testing for mbx. Effectively a typo.
7951 15. The spa authenticator server code was not checking that the input it
7952 received was valid base64.
7954 16. The debug output line for the "set" modifier in ACLs was not showing the
7955 name of the variable that was being set.
7957 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7959 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7960 lookup on its own. Something else was needed to trigger the lookup. For
7961 example, a match in host_lookup or the need for a host name in a host list.
7962 Now, if $sender_host_name is referenced and the host name has not yet been
7963 looked up, a lookup is performed. If the lookup fails, the variable remains
7964 empty, and $host_lookup_failed is set to "1".
7966 19. Added "eqi" as a case-independent comparison operator.
7968 20. The saslauthd authentication condition could segfault if neither service
7969 nor realm was specified.
7971 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7972 error message that was logged was misleading, and incoming SMTP
7973 connections were dropped. The message is now more accurate, and temporary
7974 errors are given to SMTP connections.
7976 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7977 (except RSET) with a 421 error, until QUIT is received. However, it was
7978 failing to send a response to QUIT.
7980 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7981 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7982 ACL is now run after the helo_try_verify_hosts code.
7984 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7985 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7986 case-independent, so other case variants are also recognized). Apparently
7987 some systems use these upper case variants.
7989 25. If more than two messages were waiting for the same host, and a transport
7990 filter was specified for the transport, Exim sent two messages over the
7991 same TCP/IP connection, and then failed with "socket operation on non-
7992 socket" when it tried to send the third.
7994 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7996 27. The extern definition of crypt16() in expand.c was not being excluded when
7997 the OS had its own crypt16() function.
7999 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8000 as a preferred synonym for return_size_limit, both as an option and as an
8003 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8005 30. Changed the default configuration ACL to relax the local part checking rule
8006 for addresses that are not in any local domains. For these addresses,
8007 slashes and pipe symbols are allowed within local parts, but the sequence
8008 /../ is explicitly forbidden.
8010 31. SPA server authentication was not clearing the challenge buffer before
8013 32. log_message in a "warn" ACL statement was writing to the reject log as
8014 well as to the main log, which contradicts the documentation and doesn't
8015 seem right (because no rejection is happening). So I have stopped it.
8017 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8018 However, I am unable to do any testing of this.
8020 34. Fixed an infelicity in the appendfile transport. When checking directories
8021 for a mailbox, to see if any needed to be created, it was accidentally
8022 using path names with one or more superfluous leading slashes; tracing
8023 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8025 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8026 contents are added to the log line that is written for every discarded
8027 recipient. (Previously a log_message setting was ignored.)
8029 36. The ${quote: operator now quotes the string if it is empty.
8031 37. The install script runs exim in order to find its version number. If for
8032 some reason other than non-existence or emptiness, which it checks, it
8033 could not run './exim', it was installing it with an empty version number,
8034 i.e. as "exim-". This error state is now caught, and the installation is
8037 38. An argument was missing from the function that creates an error message
8038 when Exim fails to connect to the socket for saslauthd authentication.
8039 This could cause Exim to crash, or give a corrupted message.
8041 39. Added isip, isip4, and isip6 to ${if conditions.
8043 40. The ACL variables $acl_xx are now saved with the message, and can be
8044 accessed later in routers, transports, and filters.
8046 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8047 strings in the file are not string-expanded.
8049 42. If a MAIL command specified a SIZE value that was too large to fit into an
8050 int variable, the check against message_size_limit failed. Such values are
8051 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8052 day this will have to be increased, but I don't think I want to be around
8053 when emails are that large.
8060 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8061 iconv() is not standard in FreeBSD.
8063 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8064 IPv6 enabled. The observed symptom was a segmentation fault on return from
8065 the function os_common_find_running_interfaces() in src/os.c.
8067 3. In the check_special_case() function in daemon.c I had used "errno" as an
8068 argument name, which causes warnings on some systems. This was basically a
8069 typo, since it was named "eno" in the comments!
8071 4. The code that waits for the clock to tick (at a resolution of some fraction
8072 of a second) so as to ensure message-id uniqueness was always waiting for
8073 at least one whole tick, when it could have waited for less. [This is
8074 almost certainly not relevant at current processor speeds, where it is
8075 unlikely to ever wait at all. But we try to future-proof.]
8077 5. The function that sleeps for a time interval that includes fractions of a
8078 second contained a race. It did not block SIGALRM between setting the
8079 timer, and suspending (a couple of lines later). If the interval was short
8080 and the sigsuspend() was delayed until after it had expired, the suspension
8081 never ended. On busy systems this could lead to processes getting stuck for
8084 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8085 process, before it forks any delivery processes. The open lookup caching
8086 mechanism meant that the open file or database connection was passed into
8087 the delivery process. The problem was that delivery processes always tidy
8088 up cached lookup data. This could cause a problem for the next delivery
8089 process started by the queue runner, because the external queue runner
8090 process does not know about the closure. So the next delivery process
8091 still has data in the lookup cache. In the case of a file lookup, there was
8092 no problem because closing a file descriptor in a subprocess doesn't affect
8093 the parent. However, if the lookup was caching a connection to a database,
8094 the connection was closed, and the second delivery process was likely to
8095 see errors such as "PGSQL: query failed: server closed the connection
8096 unexpectedly". The problem has been fixed by closing all cached lookups
8097 in a queue runner before running a delivery process.
8099 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8100 seem to have the "const" qualifier which it has on other OS. I've
8103 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8104 *of the same type* (client or server) with the same public name that an
8105 error should be diagnosed.
8107 9. When Exim looked up a host name for an IP address, but failed to find the
8108 original IP address when looking up the host name (a safety check), it
8109 output the message "<ip address> does not match any IP for NULL", which was
8110 confusing, to say the least. The bug was that the host name should have
8111 appeared instead of "NULL".
8113 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8114 user that is built into the binary, and the -C or -D options is used, root
8115 privilege is dropped before the configuration file is read. In addition,
8116 logging is switched to stderr instead of the normal log files. If the
8117 configuration then re-defines the Exim user, the unprivileged environment
8118 is probably not what is expected, so Exim logs a panic warning message (but
8121 However, if deliver_drop_privilege is set, the unprivileged state may well
8122 be exactly what is intended, so the warning has been cut out in that case,
8123 and Exim is allowed to try to write to its normal log files.
8129 1. smtp_return_error_details was not giving details for temporary sender
8130 or receiver verification errors.
8132 2. Diagnose a configuration error if two authenticators have the same public
8135 3. Exim used not to create the message log file for a message until the first
8136 delivery attempt. This could be confusing when incoming messages were held
8137 for policy or load reasons. The message log file is now created at the time
8138 the message is received, and an initial "Received" line is written to it.
8140 4. The automatically generated man page for command line options had a minor
8141 bug that caused no ill effects; however, a more serious problem was that
8142 the procedure for building the man page automatically didn't always
8143 operate. Consequently, release 4.20 contains an out-of-date version. This
8144 shouldn't happen again.
8146 5. When building Exim with embedded Perl support, the script that builds the
8147 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8148 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8150 6. The freeze_tell option was not being used for messages that were frozen on
8151 arrival, either by an ACL or by local_scan().
8153 7. Added the smtp_incomplete_transaction log selector.
8155 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8156 was accepting AUTH without a new EHLO.
8158 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8159 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8162 10. Logging of TCP/IP connections (when configured) now happens in the main
8163 daemon process instead of the child process, so that the TCP/IP connection
8164 count is more accurate (but it can never be perfect).
8166 11. The use of "drop" in a nested ACL was not being handled correctly in the
8167 outer ACL. Now, if condition failure induced by the nested "drop" causes
8168 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8169 or "require"), the connection is dropped.
8171 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8172 that yield "discard" can now be used with an "accept" or a "discard" verb,
8173 but an error is generated for any others (because I can't see a useful way
8174 to define what should happen).
8176 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8177 are now processed in the same way as lines in the Exim configuration file.
8178 In particular, continuation lines are supported.
8180 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8182 15. Added -ti meaning -t -i.
8184 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8185 domains, and warn by logging if others are found.
8187 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8188 returned from GIFCONF when Exim is trying to find the list of interfaces on
8189 a host. The code in os.c has been modified to copy each ifreq to an aligned
8190 structure in all cases.
8192 Also, in some cases, the returned ifreq's were being copied to a 'struct
8193 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8194 means the last couple of bytes of an IPv6 address could be chopped if the
8195 ifreq contained only a normal sockaddr (14 bytes storage).
8197 18. Named domain lists were not supported in the hosts_treat_as_local option.
8198 An entry such as +xxxx was not recognized, and was treated as a literal
8201 19. Ensure that header lines added by a DATA ACL are included in the reject log
8202 if the ACL subsequently rejects the message.
8204 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8205 MD5 (which is deprecated).
8207 21. When testing a filter file using -bf, Exim was writing a message when it
8208 took the sender from a "From " line in the message, but it was not doing so
8209 when it took $return_path from a Return-Path: header line. It now does.
8211 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8212 with a valid header line field name (a series of printing characters
8213 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8215 23. Changed "disc" in the source to "disk" to conform to the documentation and
8216 the book and for uniformity.
8218 24. Ignore Sendmail's -Ooption=value command line item.
8220 25. When execve() failed while trying to run a command in a pipe transport,
8221 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8222 could be confused with a return value of 69 from the command itself. This
8223 has been changed to 127, the value the shell returns if it is asked to run
8224 a non-existent command. The wording for the related log line suggests a
8225 non-existent command as the problem.
8227 26. If received_header_text expands to an empty string, do not add a Received:
8228 header line to the message. (Well, it adds a token one on the spool, but
8229 marks it "old" so that it doesn't get used or transmitted.)
8231 27. Installed eximstats 1.28 (addition of -nt option).
8233 28. There was no check for failure on the call to getsockname() in the daemon
8234 code. This can fail if there is a shortage of resources on the system, with
8235 ENOMEM, for example. A temporary error is now given on failure.
8237 29. Contrary to the C standard, it seems that in some environments, the
8238 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8239 program. Exim now does this explicitly; it affects the formatting of
8240 timestamps using strftime().
8242 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8243 complaints. I've now initialized all the variables, to avoid this.
8245 32. Header lines added by a system filter were not being "seen" during
8246 transport-time rewrites.
8248 33. The info_callback() function passed to OpenSSL is set up with type void
8249 (*)(SSL *, int, int), as described somewhere. However, when calling the
8250 function (actually a macro) that sets it up, the type void(*)() is
8251 expected. I've put in a cast to prevent warnings from picky compilers.
8253 34. If a DNS black list lookup found a CNAME record, but there were no A
8254 records associated with the domain it pointed at, Exim crashed.
8256 35. If a DNS black list lookup returned more than one A record, Exim ignored
8257 all but the first. It now scans all returned addresses if a particular IP
8258 value is being sought. In this situation, the contents of the
8259 $dnslist_value variable are a list of all the addresses, separated by a
8262 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8263 to accept a host name and all its aliases if the forward lookup for any of
8264 them yielded the IP address of the incoming connection. Now it accepts only
8265 those names whose forward lookup yields the correct IP address. Any other
8266 names are discarded. This closes a loophole whereby a rogue DNS
8267 administrator could create reverse DNS records to break through a
8268 wildcarded host restriction in an ACL.
8270 37. If a user filter or a system filter that ran in a subprocess used any of
8271 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8272 the wrong values were passed to the pipe command ($thisaddress had the
8273 value of $0, $0 had the value of $1, etc). This bug was introduced by
8274 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8276 38. Improved the line breaking for long SMTP error messages from ACLs.
8277 Previously, if there was no break point between 40 and 75 characters, Exim
8278 left the rest of the message alone. Two changes have been made: (a) I've
8279 reduced the minimum length to 35 characters; (b) if it can't find a break
8280 point between 35 and 75 characters, it looks ahead and uses the first one
8281 that it finds. This may give the occasional overlong line, but at least the
8282 remaining text gets split now.
8284 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8285 file descriptors might be low, and that setting 1000 would always raise it.
8286 It turns out that in some environments, the limit is already over 1000 and
8287 that lowering it causes trouble. So now Exim takes care not to decrease it.
8289 40. When delivering a message, the value of $return_path is set to $sender_
8290 address at the start of routing (routers may change the value). By an
8291 oversight, this default was not being set up when an address was tested by
8292 -bt or -bv, which affected the outcome if any router or filter referred to
8295 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8296 log only when "message" or "log_message" are set. However, if one of the
8297 conditions was an address verification, or a call to a nested ACL, the
8298 messages generated by the underlying test were being passed through. This
8299 no longer happens. The underlying message is available in $acl_verify_
8300 message for both "message" and "log_message" expansions, so it can be
8301 passed through if needed.
8303 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8304 new expansion $bh_ to give the encoded byte string without charset
8305 translation. Translation happens only if iconv() is available; HAVE_ICONV
8306 indicates this at build time. HEADERS_CHARSET gives the charset to
8307 translate to; headers_charset can change it in the configuration, and
8308 "headers charset" can change it in an individual filter file.
8310 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8311 that creates RFC 2047 encoded "words" labels them as that charset instead
8312 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8313 expansion operator; (ii) when Exim creates a From: line for a local
8314 message; (iii) when a header line is rewritten to include a "phrase" part.
8316 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8317 buggy, causing it to skip the first lines of messages whose message ID
8318 ended in 'D'. This would not have bitten before Exim release 4.14, because
8319 message IDs were unlikely to end in 'D' before then. The effect was to have
8320 incorrect size information for certain domains.
8322 45. #include "config.h" was missing at the start of the crypt16.c module. This
8323 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8326 46. If there was a timeout during a "random" callout check, Exim treated it as
8327 a failure of the random address, and carried on sending RSET and the real
8328 address. If the delay was just some slowness somewhere, the response to the
8329 original RCPT would be taken as a response to RSET and so on, causing
8330 mayhem of various kinds.
8332 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8333 when I implemented it. It didn't allow for the fact that some option values
8334 may legitimately be negative (e.g. size_addition), and it didn't even do
8335 the right test for positive values.
8337 48. Domain names in DNS records are case-independent. Exim always looks them up
8338 in lower case. Some resolvers return domain names in exactly the case they
8339 appear in the zone file, that is, they may contain uppercase letters. Not
8340 all resolvers do this - some return always lower case. Exim was treating a
8341 change of case by a resolver as a change of domain, similar to a widening
8342 of a domain abbreviation. This triggered its re-routing code and so it was
8343 trying to route what was effectively the same domain again. This normally
8344 caused routing to fail (because the router wouldn't handle the domain
8345 twice). Now Exim checks for this case specially, and just changes the
8346 casing of the domain that it ultimately uses when it transmits the message
8349 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8352 50. If a filter generated a file delivery with a non-absolute name (possible if
8353 no home directory exists for the router), the forbid_file option was not
8356 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8357 the existing equality matching.
8359 52. Exim was using ints instead of ino_t variables in some places where it was
8360 dealing with inode numbers.
8362 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8363 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8364 called TMPDIR, and if it finds it is different, it changes its value.
8366 54. The smtp_printf() function is now made available to local_scan() so
8367 additional output lines can be written before returning. There is also an
8368 smtp_fflush() function to enable the detection of a dropped connection.
8369 The variables smtp_input and smtp_batched_input are exported to
8372 55. Changed the default runtime configuration: the message "Unknown user"
8373 has been removed from the ACL, and instead placed on the localuser router,
8374 using the cannot_route_message feature. This means that any verification
8375 failures that generate their own messages won't get overridden. Similarly,
8376 the "Unrouteable address" message that was in the ACL for unverifiable
8377 relay addresses has also been removed.
8379 56. Added hosts_avoid_esmtp to the smtp transport.
8381 57. The exicyclog script was not checking for the esoteric option
8382 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8383 will work only if exicyclog is run under the appropriate euid.
8385 58. Following a discussion on the list, the rules by which Exim recognises line
8386 endings on incoming messages have been changed. The -dropcr and drop_cr
8387 options are now no-ops, retained only for backwards compatibility. The
8388 following line terminators are recognized: LF CRLF CR. However, special
8389 processing applies to CR:
8391 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8392 nor a local message in the state where . is a terminator.
8394 (ii) If a bare CR is encountered in a header line, an extra space is added
8395 after the line terminator so as not to end the header. The reasoning
8396 behind this is that bare CRs in header lines are most likely either
8397 to be mistakes, or people trying to play silly games.
8399 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8400 was being incorrectly given as 18 bytes larger than it should have been.
8401 This is a VOB (very old bug).
8403 60. This may never have affected anything current, but just in case it has:
8404 When the local host is found other than at the start of a list of hosts,
8405 the local host, those with the same MX, and any that follow, are discarded.
8406 When the list in question was part of a longer list of hosts, the following
8407 hosts (not currently being processed) were also being discarded. This no
8408 longer happens. I'm not sure if this situation could ever has previously
8411 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8414 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8415 header. This is something that is recommended in a new Internet Draft, and
8416 is something that is documented as being done by Sendmail. There are two
8417 possible values. For messages generated by the autoreply transport, Exim
8420 Auto-Submitted: auto-replied
8422 whereas for all other generated messages (e.g. bounces) it adds
8424 Auto-Submitted: auto-generated
8426 63. The "personal" condition in filters now includes a test for the
8427 Auto-Submitted: header. If it contains the string "auto-" the message it
8428 not considered personal.
8430 64. Added rcpt_include_affixes as a generic transport option.
8432 65. Added queue_only_override (default true).
8434 66. Added the syslog_duplication option.
8436 67. If what should have been the first header line of a message consisted of
8437 a space followed by a colon, Exim was mis-interpreting it as a header line.
8438 It isn't of course - it is syntactically invalid and should therefore be
8439 treated as the start of the message body. The misbehaviour could have
8440 caused a number of strange effects, including loss of data in subsequent
8441 header lines, and spool format errors.
8443 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8444 client host had authenticated. This control can now be exercised by an ACL
8445 for more flexibility.
8447 69. By default, callouts do not happen when testing with -bh. There is now a
8448 variant, -bhc, which does actually run the callout code, including
8449 consulting and updating the callout cache.
8451 70. Added support for saslauthd authentication, courtesy of Alexander
8454 71. If statvfs() failed on the spool or log directories while checking their
8455 size for availability, Exim confusingly gave the error "space shortage".
8456 Furthermore, in debugging mode it crashed with a floating point exception.
8457 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8458 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8459 serious problem, Exim now writes to the main and panic logs when this
8460 happens, with details of the failure. It then refuses to accept the
8461 incoming message, giving the message "spool directory problem" or "log
8462 directory problem" with a 421 code for SMTP messages.
8464 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8465 0, 1, and 2 exist, because some OS complain for execs without them (see
8466 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8467 descriptors. However, the code omitted to check that the open succeeded,
8468 causing mysterious errors if for some reason the permissions on /dev/null
8469 got screwed. Now Exim writes a message to the main and panic logs, and
8470 bombs out if it can't open /dev/null.
8472 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8473 interact so that it is all more flexible. It is supposed to remain
8474 backwards compatible. Also added extra_local_interfaces.
8476 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8477 to bomb out with an assertion failure - to the client this appears as a
8478 connection drop. This problem occurs in the part of the code that was taken
8479 from the Samba project. Fortunately, the assertion is in a very simple
8480 function, so I have fixed this by reproducing the function inline in the
8481 one place where it is called, and arranging for authentication to fail
8482 instead of killing the process with assert().
8484 75. The SPA client code was not working when the server requested OEM rather
8485 than Unicode encoding.
8487 76. Added code to make require_files with a specific uid setting more usable in
8488 the case where statting the file as root fails - usually a non-root-mounted
8489 NFS file system. When this happens and the failure is EACCES, Exim now
8490 forks a subprocess and does the per-uid checking as the relevant uid.
8492 77. Added process_log_path.
8494 78. If log_file_path was not explicitly set, a setting of check_log_space or
8495 check_log_inodes was ignored.
8497 79. If a space check for the spool or log partitions fails, the incident is now
8498 logged. Of course, in the latter case the data may get lost...
8500 80. Added the %p formatting code to string_format() so that it can be used to
8501 print addresses in debug_print(). Adjusted all the address printing in the
8502 debugging in store.c to use %p rather than %d.
8504 81. There was a concern that a line of code in smtp_in.c could overflow a
8505 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8506 initially expressed, the concern was not well-founded, because trailing
8507 spaces are removed early. However, if the trailing spaces were followed by
8508 a NULL, they did not get removed, so the overflow was possible. Two fixes
8511 (a) I re-wrote the offending code in a cleaner fashion.
8512 (b) If an incoming SMTP command contains a NULL character, it is rejected
8515 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8516 runs initgroups(), so that if the Exim user is in any additional groups,
8517 they will be used during message reception.
8523 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob